Skip to main content

Full text of "Bhartrihari's Vakyapadiya"

See other formats


1. The Brahman who is without beginning or end, 
whose very essence is the Word, who is the cause of 
the manifested phonemes, who appears as the objects, 
from whom the creation of the world proceeds, 

It 1 is solemnly declared here that Brahman is beyond all 
representation ; 2 it is endowed with all powers which are 
neither identical with it nor different from it; it has two 
aspects, that of unity ( vidya ) and that of diversity ( avidyd ); 
(even in its aspect of avidyd ), it is really free from all diver¬ 
sity; it is, in all states, unaffected by beginning and end, even 
though the manifestations appear in wordly transactions 3 in 
a temporal and spatial sequence. The Brahman is both effect 
and cause, it is many and one and in all the different systems, 
the manifestations are not conceived as having nothing before 
them and nothing after them. Nor is any limit admitted, above 
below, or sideways, to its spatial differentiation. All its mani¬ 
festations, though they appear to be distinct from one another, 
are in the nature of the word ( sabda ), because in all of 
them, the original material persists. Because, in our cog¬ 
nitions, we identify objects with their words (sabdopagrdhi) 
and our cognitions are intertwined with the words (sabdo- 
pagrdhya), they are essentially of the nature of the word . 4 
All divisions such as the beginning, the sustenance and the 
end of things can be determined only through words. The 
Brahman is called phoneme ( aksara ) because it is the cause 
of the phonemes . 5 The manifestation of the word which is 
in everybody as one with the spirit is for the purpose of 
communicating what is within. It has been said: — 

“Some declare that the Word which is really subtle 
identical with the meanning, One and identical with the 
Self when not manifest, appears as something else and diffe¬ 


What is meant by ‘it appears as the objects’ is this—what 
is called appearance (vivarta) is the assumption by the One, 
without losing its one-ness, through apparent diversity, of 
the unreal forms of others.6 It is like the appearances in a 
dream. It has been said—“The spatial and temporal mani¬ 
festations are no more than the functioning of the powers of 
Nescience they are neither identical nor different from the 
one which is the Truth. The fact of being Nescience consists 
just in that What is meant by “from whom the creation of 
the world proceeds” is this:—The transformations, that is, the 
Universe proceeds out of Brahman which is the Word, devoid 
of all inner sequence; from that involution (smnvarta) in 
which all diversity has merged and is undifferentiated and is 
inexpressible, all transformations being in a latent stage. 7 

/ a \8 The Self which, though all representations are of it, does not 
come within their range and which, on the basis of guess, authority 
and inference, appears in different ways. 

(bl The all-embracing, beyond all differentiation and unification, 
existence and non-existence, sequence and absence of it, truth and 
falsehood, shines as distinct from everything. 

He the inner Controller of Beings, is seen near and far. 
He i uSly f-e and is sought by those who desire liberation. 

(d) He is the creator of those objects which are looked upon as 
ultimate causes, just as the lustre of the seasons is the creator of the 
masses of clouds at the end of summer. 

(el That one spirit is differentiated in many ways, like the waters 
of the ocean, which are impregnated with heat at the time of dissolution. 

(f) From that spirit which is like a universal and liable to be 
differentiated, legions of particulars are produced, like ram-laden clouds 
from special winds. 

(g) That supreme light appears first as the three Vedas and is 
the cause of different views among the followers of different systems. 

(h) That aspect of it which is really the truth, full of peace, is, 
however, always accompanied by Nescience, which is indefinable.9 

(i) There is no limit to the number of transformations of it (Nes¬ 
cience). When the soul has realised it, it does not exist for him any 

(j) Just as one having a defect in the eye sees the clear sky as 
being clouded with many forms, in the same way, the immortal 

1 i. 3 

Brahman, free from all change, is soiled by Nescience and appears to 
have many forms. 

(k) This Brahman creates everything as having the nature of the 
word; it is the source of the illumination power of the word. This 
universe emerges out of the word aspect of Brahman and merges, into 

[1. Sometimes, where the text consists of a long sentence, the 
translation is in the form of short phrases. It is hoped that it will be 
clear to the reader of which part of the long sentence each short 
phrase is a translation. 

2. The word so translated is parikalpa. An attempt will be made 
to use the same English word for translating a Sanskrit word occurring 
in different contexts; but in some cases, other more suitable English 
words will be found to have been used. 

3. Vyavahdra. This word occurs frequently. Here it has been trans¬ 
lated as ‘worldly transactions’. Often it means the use of words to 
communicate one’s thoughts, verbal usage, or exchange of ideas. As 
Helaraja puts it— pratipadyapratipddakayor hi parasparabhiprayasah- 
krdxitir vyavaharah (He. on Vak. III. Sain. 32) . On another occasion, 
he says— Jhanasyaiva parcisparasahkrdntir vyavaharah (He. on Vak. III. 
Sain. 2), prakhyopdkliyatmakatvacca vyavaharasya. . . (He. on Vak. III. 
Dra. 14) . 

4. vikdrarvarri prakrtyanvayitvac cliabdopagrahyataya sabdopagrdhi- 
taya ca sabdatattvam. This is one of the important arguments in the 
chain of reasoning meant to prove sabdadvaita, the doctrine that the 
ultimate reality from which everything comes is in the nature *of word. 
The argument is used in the different systems of philosophy also. 

5. aksara ordinarily means ‘imperishable’, but the Vrtti gives a 
special meaning to the word. 

6. The Vrtti gives here a defination of the important word vivarta , 
in connection with the explanation of the word vivartate. According 
to the Vrtti, the word, though a verb, also conveys the same idea. 

7. purvam vikaragranthirupatvenavyapadesyat (Vr). Even with the 
help of the Paddhati, the meaning is not clear. The word granthi occurs 
several times in the text, both in the Karika and in the Vrtti:—Vak. I. 
115; Vrtti on Vak. I, 4, 5, 142; 145. Usually it has the meaning of trans¬ 
formation ( vikara ) but here it comes in the same compound with the 
word vikara and so it must have a different meaning. As Vrsabha puts 

it _ granthisabdo’ nyatra yadyapi vikaravacanah tathapi vikdrasabdena 

salia pryujyamdno’ peksitapravrttinimitto drastavyah (Vr. on Vak. I. 1. 
(p. 10, 1. 14). Here the whole expression in question qualifies brah- 
manah. The translation is very tentative. 

8. This and the following verses are called agama by Vrsabha and, 
according to him, they elaborate the idea of the Karika which is being 



explained. As he puts it —iddnim sarvasyaiva kdrikarthasyagamendnuga - 
mam darsayati. (Vr. on Vak. 1.1. (p. 10, 1. 17). 

9. santavidyatmakam etc. The construction of this verse is defec¬ 
tive. In view of amsah, one would expect dtmakah and grastah , but all 
the mss of the Vrtti and the Paddhati have the neuter form). 

2. Who has been taught as the One appearing as 
many due to the multiplicity of his powers, who, though, 
not different from his powers, seems to be so, 1 

It has been revealed that the diversity of the transfor¬ 
mations and the unity of what is transformed do not go be¬ 
yond the unity of the Ultimate (prakrtyekatva) , 2 As has 
been said—‘It is like Water, the Seer, One without a second * 
Similarly: ‘O gentle One!, in the beginning, Being alone was 
One without a second’ * Then again— ‘The one Pranava was 
divided into three’. Similarly, ‘In the beginning, there was 
non-being. What was that non-being? It is the Rsis who were 
non-being at the beginning, the Rsis who were' the prcinas ”5 
What is meant by ‘due to the multiplicity of his powers’ is— 
Appearing as many, the powers which are mutually opposed 
and are identical with Brahman accumulate in it which is 
essentially the Word. In a cognition in which many objects 
figure, the different objects which figure such as earth, people, 
etc., do not affect the unity of the cognition. There is no con¬ 
tradiction between the multiplicity of the things like trees 
which are cognised and the unity of the cognition. The form 
of the cognition does not really differ from that of the object 
because different forms of the objects are not beyond the 
unity of the cognitions. Similarly, the powers which appear 
to be different from one another are not really so. The text 
‘though not different from his powers’ means: the powers are 
not different from Brahman as the universal and the particu¬ 
lar are from each other. But it appears to be different when 
it assumes the form of the different objects which figure in it. 6 

[1. Bhinnam saktivyapasrayat’ would perhaps be a better read 
ing but all the mss and the Paddhati have the text as printed. 

2. prakrtyekatvanatikramena. By ‘oneness’ of Brahman only ab 
sence of all differentiation-is meant and not association with the number 

I. 3. 5 

one. As Vrsabha says— na tvekatvasayikhyayogena. ‘Oneness’ due to 
association with the number one belongs to manifestations and not to 
the ultimate. There are two kinds of oneness, says Vrsabha: dvidham 
ekatvam prdkrtam vaikrtam ca (Vr on Vak. I. 1. (p. 15, 1 . 2). 

3. Cf. Br. Up. 4.3.32 where the text is— salila, eko drasta? dvuito 


4. Ch. Up. 6.2.1. 

5. 6. Br. 

6. The word sakti is used twice in this verse. Vrsabha explains it— 
ghatddayali paddrthah yogyatdva, (Vak. I. 2. (p. 14, 1. 20). in other 
words, both the power to bring about appearances and the appearances 
themselves can be called sakti]. 

3. Depending on whose Time-power to which 
(though one) differentiation is attributed, the six trans¬ 
formations, birth etc. become the cause of all variety in 

All powers 1 depending on their causes and having a 
starting point are governed by the creative power ( svdtan- 
trya.) called Time; they follow the modes of this Time-power. 
Because of the regulation of the immense diversity of each 
object by means of prevention and permission, its appearances 
seem to have a sequence. Time, by allowing them to come to 
be, is the secondary cause of all transformations which de¬ 
pend on their own other causes also and whose production 
had been delayed. 2 Its causal power being thus differentiated 
by the manifestations which have a sequence, the differentia¬ 
tion which exists in the manifestations is attributed to it. It 
is like attributing to the balance the divisions marked by lines 
on it at the time of the balancing of the weight of the mate¬ 
rial which is in contact with it. 3 Thus when sequence is 
attributed to something which is neither prior nor posterior 
in the form ‘it was’ or ‘it was not,’ the six transformations 
birth etc., become the source of the modifications, the 
changes in Being. Transformation of action will be dealt 
with in detail in the section devoted to ‘Being’ under the 
subject of universal. 



[1. Vrsabha explains saktayali as padarthah by identifying cause and 

2. pratibaddhajanmandm. Though all the mss of the Vrtti have 
pratibandha° the text adopted is probably the genuine one. The ex¬ 
planation of Vrsabha— yesdmevdTiena pvathdTiidta utpattau pvatibandhuli 
krtah te pratibaddhajanmdnah’ also supports the emendation. 

3. While explaining the analogy of the balance, Vrsabha uses the 
terms tuldsutTd, tulddanda, tulupdtdla and ddiulalekhd, The rod type of 
balance seems to be referred to. The thing to be weighed is placed 
on a pan (tuldpatald) which hangs from one end of the rod (tuld- 
danda) which is held by a string ( tuUsutra ) at one of the points 
marked by a line on the rod ( ddiiddlekha ) according to the weight of 
the thing weighed]. 

4. Of which one that is the seed of all, there is 
this state of multiplicity, that of the enjoyer, the enjoyed 
and enjoyment. 1 . . , 

The One Brahman is the substratum of powers which 
cannot be defined as identical or different from it, as exis¬ 
tent or non-existent and are not opposed to one another; 2 it 
is differentiated into unreal forms which, like the beings 
appearing in a dream, have no external reality, are different 
from one another and are knots of the nature of enjoyer, 
enjoyed and enjoyment. When the predispositions towards 
differentiation mature, this worldly arrangement with its 
multiplicity comes into being. 3 

[1. The main purpose of this verse is to point out that nothing is 
different from Brahman and that, yet, all appearance of differentiation 
is due to it. 

2. Even though the powers produce opposite effects and are, in 
that sense, opposed to one another, yet they are said to be not so, 
because they all exist in the same substratum at the same time. 
(ekasminn adhdne yaugapadyena vrtteli —Vr.). 

3. tasya granthyantarasamatikramena vivrttagranthiparicchedasya. 
This is rather obscure. The word granthi comes twice. See note, 7 on 
verse 1. The commentary ‘Ambakartrl’ explains granthi as vasand, the 
predispositions leading to differentiation, whereas Vrsabha takes it to 
mean the differentiated objects themselves. Something which appears as 
a differentiation within something which has no differentiation is a 

I. 5. 


grantlii. As Vr. puts it, ata eva pavanuitinano vihtircitvHd granthaydh. I 
have used the word ‘knot’ suggesting a distortion of reality. The trans¬ 
lation is very tentative]. 

5. A means of attainment and a symbol of that 
One is the Veda, which though one, has been handed 
down as though in many recensions by the sages. 

The attainment of Brahman is nothing more than going 
beyond the knot 1 of the ego-sense in the form of ‘I’ and ‘Mine’. 
Others say that it is the resolution of the effects into the 
original cause, 2 the cessation of the activity of the senses, 3 
satisfaction without the aid of external means, one’s own self, 4 
the yearning for the supreme Self, the absence of craving for 
any adventitious object, the fact of possessing complete 
power, 5 escape from the influence of the functions of Time,® 
the complete cessation of the self. 7 Such are the alternative 
ways of conceiving the attainment of Brahman. The means of 
attainment is the collection of the Vedas. Just as making 
gifts, pel-forming austerities and practising continence are 
means of attaining heaven. It has been said: When, by prac¬ 
tising the Vedas, the vast darkness is removed, that supreme, 
bright, imperishable light comes into being in this very 
birth. By the word symbol (anuhara) the idea contained in 
the following ancient saying (purdkalpa) is meant: The Rsis 
the seers of the mantras, those who have realised the truth 
(dharma) see that subtle, inaudible Word and, wishing to 
communicate it to those who have not realised the truth, 
teach the symbol of it which is like a dream 8 in their desire 
to tell what they have seen, heard and experienced. It has, 
indeed, been said: — The Rsis realised that truth (dharma ); 
they taught the mantras to those who had not realised the 
truth- these others, also anxious to teach, proclaimed the 
Vedas and the Vedangas, in order that the symbol of Brah¬ 
man may be understood (bilmagrahandya ). Bilma is bhilma 
which means something which illuminates (bhdsanam) . 9 By 
the words ‘The Veda, though One, has been handed down in 
many ways by the Rsis,’ what is meant is this; The object 


vakyapadiyam of bhartrhari 

called Veda, as seen by the Rsis in their vision, is One. As 
the One cannot be communicated through the many, the 
Rsis first transferred it to the Word to which diversity is attri¬ 
buted because of the diversity of the manifesting agents (i.e., 
the dhvanis). The Word, then, without giving up its unity, 
assumes different forms such as Samhitd , Pada, Krama etc. 
It is handed down by the Rsis in different branches, establish¬ 
ed for the sake of convenience of study. Another view is: 
The Word varies in the different regions and has, therefore, 
many forms; but it does not deviate from its purpose of con¬ 
veying the same thing everywhere. In its different forms, it 
becomes the cause of regional and other diversity. In the 
same way, the words of the Vedas, though handed down in 
different branches, do not deviate from the fact that every¬ 
where they convey the same thing. The diversity in the 
branches is based on the diversity of forms. Others are of 
this view: In the old days, the Ayurveda, with its eight 
sections, 10 was one. In the Kali era, due to the reduced 
capacity of men, it has been divided into sections. In the 
same way, the Veda, the collection of Brahman, includes an 
infinite number of ways and powers. 

[1. Granthi. See n. 7 on verse 1 and n. 3 on verse 4. 

2. Ambakartri understands this as the Sankhya view. 

3. Ambakartri sees here the yoga view. The word vikaranablidva 
occurs in Yog. Dar. 3.48 in connection with the attainment of super¬ 
natural powers. 

4. dtmatattva. While explaining this, Vr. gives reference to Br. 
Up. 4.3.21. where the text, in a slightly modified form, is as follows: 
tad yatlvd priyaya striya samparisvakto na baliyam kincana veda. The 
idea is that when Brahman is attained one is aware of nothing else. 

5. paripurnasaktitvam. Vr. explains sakti in the sense of the eight 
supernatural powers attained through yoga. They are: anima, mahima, 
garimd, laghima , prdpti , prdkamya, isitvam, vasitvam. See Y. Dar. 3.45. 

6. Kalavrtti. The two functions of time, prevention (pratibandha) 
and permission ( abhyanujna ) by virtue of which only certain effects 
appear at certain times and not others. 

7. sarvatmana nairdtmyam. Vr. understands this as referring to 
the extreme indefinability of Brahman: nailisvabhavyam, brahmasva- 
bhavasya nirupayitum vaktum cdsakyatvat. Some scholars, however, 
take it as referring to the Bauddha view. 


I. 6. 

8. Svapnavrttam iva. Just as one’s experience in a dream is a kind 
of reflection of our experience in the wakeful state, in the same way, 
the Vedas are a kind of reflection of what the R$is saw in their vision. 

9. Ni. 1.20.2. 

10. The eight sections of Ayurveda, as given by Vr. are: salyacikitsa, 
salakyam, kayacikitsa 3 bhutacikitsa , kaumarabhrtyam, agadatantmm, 
vajzkaranatantram, rasayanatantram ]. 

6. Its divisions follow many recensions, but they 
are all subsidiary to the same action. Also a certain 
fixity in the power of words is seen in these different 

Once the Veda has been divided into four, there are 
the one hundred branches of the adhvaryus ( Yajurveda ), 
the thousand paths of the Samaveda, the twenty-one kinds of 
Rg-Veda or fifteen as some say and the nine kinds of Athar- 
vana veda. Thus every Veda has many paths. What is meant 
by ‘subsidiary to the same action’ is: All branches lead to the 
same ritual. It is like this: all the branches of the physicians 
ultimately practice one treatment. What is meant by ‘fixity 
in the power of words’ is: Words, as, taught for a particular 
branch, convey their meaning in that branch and also lead 
to merit. What is meant by ‘it is seen in the different bran¬ 
ches of Veda’ is seen in the following examples “Long d 
is substituted for the final of deva and sumna when the 
suffix kyac follows in the Kathaka branch of the Yajurveda.” 
(P. 7, 4, 38.) and ‘the final of the word sima bears the acute 
(udatta) accent in the Atharvana Veda’ (Phitsutra 79.) All 
this is according to the view of those who hold that the divi¬ 
sion of the Veda into branches takes place again and again 
and that, before the division, the word invariably exists in 
an undifferentiated state. 1 

[1. Vr. points out that what has been said in the verses 5 and 6 
is based on the view that the Veda manifests itself as one from Brahman. 
Later, due to the reduced capacity of men, it is divided into branches. 
At the time of dissolution ( pralaya ) it again becomes one. When crea- 
K. S. 2 



tion begins again, it manifests itself as one but it is again divided into 
branches when the capacity of man diminishes. 

The other view is that the Veda manifests itself from the very 
beginning with divisions. According to this, there is no undifferentiated 
word. As Vr. puts it: tesam upasamhrtakramd vdg durlabha. (Vak. 
1.6. (p. 27, 1. 24).] 

7. The various Smrtis, some having visible utility 
and others invisible utility, have been propounded by 
the sages, well-versed in the Veda on the basis of this 
very Veda, with the help of indications. 

Some traditions have a written basis while others have 
no written basis, but are known from the conduct of the 
cultured. Traditions relating to medical treatment etc., have 
a visible purpose to serve. Traditions relating to what can 
be eaten and what not, which woman one can marry, or 
otherwise, what can be said and what not have an invisible 
purpose. Where two traditions relating to the same subject 
go against each other, there is option, provided that there is 
no visible purpose to be served. 1 Where two traditions rela¬ 
ting to the same subject go against each other and there is 
a visible purpose and disapproval by the cultured, there, the 
traditions having visible purpose have no authority. 2 Where 
two traditions relating to the same subject differ from each 
other and there is visible purpose, but no disapproval by the 
cultured, there there is option. For example, in the matter 
of the penance to be observed for killing a frog. 3 What is 
meant by ‘on the basis of this very Veda, with the help of 
indications/ is that the indications found in the Scripture 
make us know that the actions taught in the Scripture and 
the written Tradition have the same agent. Those persons 
who are entitled to perform the action taught by Scripture, 
having a visible or invisible purpose, are made known by 
the Scripture itself as being entitled to perform the actions 
taught by the written Tradition. For example, the injunc¬ 
tion: ‘the sacrificer should cook for a guest who has arrived 
a big bull or a big ram.’ 4 Sometimes (i.e., when no indic$- 

i. s. ii 

tion is found in the Scripture in a particular case) the indi¬ 
cations found elsewhere are enough on the analogy of the 
rice in the cooking pot, 5 to establish the authority of the 
traditions which do not go against Scripture. 

[1. G.Dh.S. 22, 3 and Manu. 11. 73 prescribe two different penan¬ 
ces for one who is guilty of having killed a Brahmana. As the purpose 
of the penance is invisible and as both texts are authority, there is 
option. Another example is G.Dh.S. 23, 8-10 where two penances are 
taught for one who is guilty of having committed sacrilege with the 
wife of the Guru. 

2. Even where a tradition has a visible purpose to fulfil, if it is 
disapproved by the cultured, it should not be followed. Drinking of 
liquor as a medicine has a visible purpose, namely, the curing of the 
disease and it is, therefore, prescribed by Ayurveda. But drinking 
liquor is prohibited in the Dharmasastra i.e., by the cultured. This 
prohibition has greater authority than the prescription of Ayurveda. That 
is why one has to do penance after drinking liquor as a medicine. 

3. Ya. Smr. 3. 270. 

4. &. B. 3. 4. 1. 2. 

5. See Jacob. A Iiandjul of Popular Maxims, I, p. 52]. 

8. Based on the explanatory comments and simi¬ 
lar passages, conflicting views have been set forth by 
the exponents of Monism and Dualism according to 
their own taste. 

It is generally seen that human speculations are based 
upon explanatory comments and similar Vedic sentences. 1 
Because of the diversity of the human intellect, diversity 
of speculation takes place. For example, there is the expla¬ 
natory comment ( arthavMa ): “In the beginning, all this 
was nothing,” meant to praise the place where the fire is to 
be kindled. 2 On the basis of an imaginary interpretation of 
it, the following Absolutism is constructed: ‘The non-exis¬ 
tent is born of the non-existent, the action-less from the 
action-less, the non-substratum from the non-substratum, the 
inexpressible from the inexpressible, the essenceless from 
the essenceless.’ 



“The One Self is differentiated by means of existence 
and non-existence, neither of which is prior or posterior. 
How can the undifferentiated One be both expressible and 
inexpressible ?” 


“There was no non-existent nor was there the existent.” 3 

“In the beginning, all this was only darkness.” 4 

“This (the world) is like foam, it is nothing, it is like 
bubble, it is nothing. It is mere illusion, not easily set aside. 
So do the learned see it.” 

“The blind one saw the jewel, the finger-less one thread¬ 
ed it, the neck-less one wore it and the tongue-less one 
praised it.” 5 

The sentence 

“In the beginning, all this, was water”. 6 

is an explanatory comment relating to the new and full 
moon sacrifice. From that arises the following speculation: — 

“Of these waters, there is the effect (paka) in the form 
of consciousness and there is the effect in the form of matter. 
And life (pinna) is the essence of the waters and that comes 
back and also does not come back.” 7 

Really speaking, it is the one having the form of all 
knowledge and appears in all this diversity. Dualists also 
have their views: — 

“There are the eternal things and the non-eternal things, 
the causes of all objects, things which have a form and 
things which have no form, subtle and .gross, in which the 


“Two birds, companions, united together, occupy the 
same tree. Of the two, one eats the tasty fig fruit, the other 
one does not eat, but shines.” 8 


i. fi. 

“By ‘two birds’, the senses and the inner Controller, the 
intelligence and the soul are meant.” Others, 9 on the other 
hand, say: — 

“The differentiated and the undifferentiated conscious¬ 
ness, full of seeds, are meant.” 

“It moves, it does not move, it is far, ‘it is near, it is 
within everything and so is it without.” 10 

[1. The Vrtti explains artliavadarupani in the verse as arthavaddn 
arthavddaprakdi'dni. Thus the word rupdni is explained as arthavada - 
prakdrani . For Vr., the word artliavadarupani in the verse is an ekasesa; 
artliavddan artliavadarupani cety artliavadarupani on the analogy of the 
ekasesa in P. 7. 4. 82. guno yaiilukoh which means: yani yaiiluki ceti 

2. Vr. quotes the following prescription — 

Sa vai saptapuruso va 

He adds: catvara atnvdnali, trayali paksapucchdmti. te ca purusd pranaU 
vena stuyante. Cf. 6. Br. 6. 1. 1. 6. 

3. RV. 10. 129. 1. Vr. calls this passage also an arthavada . The 
idea is that by giving the place of kindling the fire the same attributes 
as Brahman, one is praising it. 

4. Cf. Mai. Up. 5. 2. where the text is: Tamo vd idam agra dsid 

5. Tai. Ar. 1. 11. 5. runs as follows — 

Sa tam manim avindat, 

So’ nangulir dvayat; 

So 3 grivali pratyamuncat 
So’ jihvo asascata. 

6. &. B. See also Jai. Up. 1.56.1. 

7. Vr explains avartti and andvartti as ‘coming back’ in the case 
of those who have not yet attained ‘ moksa 3 and ‘not coming back’ in 
the case of those who have attained it. 

8. Cf. RV. 1. 164. 20 and Sve. Up. 4.6. where the text is: 

Dvd suparua sayuja sakhaya samanam 
vrk?apri parisasvajdte . 
tayor anyah pippalam svadvatti 
anasnann any o’ bhicakasiti. 

9. According to Vr, these are again the Monists who hold this 

10. SYV. 40. 5.] 



Since there are these different speculations in regard to 
the one Brahman beyond all differentiation on account of its 
being endowed with all powers, therefore it is that 

9. The true and pure essence has also been taught 
there, the knowledge capable of being grasped through 
one word, having the form of Pranava, and in no way 
contradicting the different views. 

No representation in regard to the one all-embracing 
Brahman is different from the other representations consider¬ 
ed to be opposed to it. 1 Knowers of Brahman have declared: — 

“Even a particular manifestation of Brahman, is not be¬ 
yond its all-comprehensiveness nor is it different from 
another.” 2 


“All vision is complete and is really not different (from 
the others) but the experiencer looks upon it as different.” 3 
The mystic syllable ( Pranava ) allows for all points of view, 
it is the source of all Scripture, 4 it is the common factor of 
all original causes, it is the cause of the rise and fall of doc¬ 
trines, it accepts within itself all mutually contradictory ideas 
of Brahman or disallows all of them. The object of this allow¬ 
ance and disallowance does not, therefore, vary. 

It has been said: — 

“It is one and it is not one; it is both and it is not both. 
Those who are for action speak of diversity ( visama ) and 
those who are for knowledge see unity ( sama ).” 

[1. According to Vr. as all representations are about the same 
subject, so they are not opposed to one another: sarv&sam visaya- 
matrandm satyatvad brahmaiva visaya iti visayablied-dt parikalpantarcini 
na bhidyante. 

2. This is given as a statement of Brahmavidah but it is not found 
in the old Upanisads. 

I. 10. 


3. Vr explains nnyunam as follows — Darsandntardnmkrtaih 
paddrthair darscnvantaram api yuktam, sarvesdyi sarvarupatvdt The id ' 
seems to be that, as all systems are speculations in regard to Brahman 
which includes everything, they include one another. 

4. Sarva&rutirupali. Pranava is the source of all Scripture (sruti) 
as the Vedas are manifestations of it. Yatah pranavo Vedarupena 
vivrttah. (Vr.). It is also the source of all words (sruti): sarvasruti- 
rupali sarvasabdasvabhavah sarvasabdarupena tasya vivrtteh (Vr) ] 

10. All the different sciences which impart know¬ 
ledge and bring culture to man proceed from the major 
and minor limbs of that Veda, the creator and organiser 
of the worlds. 

The Veda, as the source and teacher of the world, is 
the creator in regard to the appearances and organiser in re¬ 
gard to the organisations. Some say that Pranava is the Veda. 
It is the source of all words and all things. According to this 
view, the different sciences, being essentially elaborations of 
the Pranava, do not go beyond the Veda. 

It has been said: — 

“All the words are included in the Veda. 

One who does not know the Veda cannot understand 
Brahman at all.” 

It has also been said: 

“The injunctions, what is enjoined and the reasoning 
constitute the Veda.” 1 

From the major ( anga ) and the minor (upaniga) limbs of the 
Pranava, from the Scripture, written Tradition and the final 
portion of the three-fold Veda are derived the different scien¬ 
ces which are the cause of right knowledge and the culture 
of man. 2 Or they may be looked upon as the cause of the 
culture of man because man consists essentially of knowledge 
(jnana) , 3 From the major auxiliary sciences like astrono¬ 
my-astrology etc., of the well-known Brahman called Veda 
come the science of omens etc., and from the minor auxiliary 
sciences of the same Veda proceed dream-lore etc, 



[1. To the question: What is the Veda?, two answers are recorded 
in the Vrtti: (1) Pranava eva veda ityeke, (2) Vidhir vidheyas tar - 
kasca vedah. 

2. Pranava } anga and updnga. The Vrtti understands these terms 
as standing for sruti, smrti and trayyanta, i.e. the explanatory com¬ 
ments ( arthavada ), the auxiliary sciences, Grammar etc., and the 
Upanisads. According to another view, sruti is pranava, smrti, is either 
anga like Grammar or updnga like Ayurveda and trayyanta is the 
Upanisads. Usually, the word anga stands for the six auxiliary scien¬ 
ces of the Veda: Jyotisa, giksa, Kalpa, Vydkarana, Nirukta and 
Chandas. From them, the science of omens ( sakunajnana ) etc are 
derived. The word updnga stands for smrtis, purana, etc., from which 
are derived dreamlore, etc. 

3. Jndnasamskdrahetavah in the verse is explained in two wa 

in the Vrtti: (1) samyagjndnahetavali purusasamskarahetavasca' 

(2) jnandtmakatvad vd purusasyaiva samskdrahetavali. This is not the 
only place where the Vrtti contains alternative explanations of the text 
of the verse. The questions which this raises will be discussed 

11. The best of all the austerities, the one that i s 
nearest to that Brahman is the discipline called 
‘Grammar’, the first among the auxiliary sciences of the 
Vedas, so have the sages declared. 

That auxiliary science through which the form of the 
Word—Brahman 1 is determined in order that correctness may 
be understood is the nearest (to the Vedas) because it helps 
directly. Proximity is known to depend upon special service 
rendered. There is also the Scripture to be restored. 2 Once 
it has been settled by the reasoning (of the Mimamsa) what is 
to be restored and what not, the science of Grammar is the 
basis for the correct modification of gender, number etc It 
is said to be near because it is through it that the other scrip¬ 
ture (which is not available) is understood. What is meant 
by ‘the best of all austerities’ is this: Compared to what are 
considered to be austerities in the world like continence, sleep¬ 
ing on the ground, standing in water and progressive reduc¬ 
tion and increase of nourishment according to the course of 
the moon 3 and compared to what are considered to be thp 

I. 12. 


holiest forms of study, the study of the science of Grammar is 
the best form of austerity, because it leads to special results, 
visible and invisible. Concerning which the tradition has 
declared that a mere knowledge of whose enumeration of 
phonemes ( aksarasamdmndya ) leads to the same holy fruit 
as the knowledge of the whole of the Veda,4 that, indeed, is 
the first auxiliary science of the Vedas, because it is the most 
important one. It has been said: — 

“Of the six auxiliary sciences, the most important one 
is the Science of Grammar. Effort made in regard to the 
most important one surely yields fruit.” 5 

[!• 17? explains cnicili as: sabdctTupBiici vivTttdsyn 


2. Uliyam camn&yantaram asti. See Mi. Su. 2.1.34. 

3. Cf. Brahmacaryam satyavacanam savanesiidakopasparsana- 

mardravastrata-dhalis&yitanasaka iti tapamsi, G.Dh.S. 3.1.15, 

4. M.Bha. I. p. 36. 

5. M.Bha. I. p. 1.] 

12. This discipline is the shortest route to the 
attainment of that supreme essence of the Word which 
has assumed differentiation, of the holiest of all lights. 

What is meant by ‘Of the word which has assumed diffe¬ 
rentiation’ is: Of the word which has been differentiated into 
phoneme, word and sentence on the basis of the place of 
articulation etc., from the undifferentiated sequenceless inner 
word-essence. Or ‘of the word which has assumed different 
forms like cow etc., as meanings on the basis of the eternal 
relation between word and meaning. Or it might be said 
that the divisions of the word assume the form of cow etc. 
Or rather, cow etc., apparent external divisions, are trans¬ 
formed into words. Some hold that only causality is the re¬ 
lation between word and meaning. 1 It has been said:_ 

“It is the name which appeared as the form and it is the 
form which appeared as the name. According to some, they 
ICS. 3 



were one at first and were differentiated later. According to 
others, the differentiation was already theie before. 

What is meant by ‘the supreme essence of the word’ is: the 
collection of words, the correctness of which has been estab¬ 
lished, both because of its expressiveness and its capacity to 
bring about happiness, is spoken of. It has been said: “words 
devoid of correctness are just the sediment.” What is meant 
by ‘the holiest of all lights’ is: In this world, there are three 
lights, three lamps which illuminate themselves and others. 
They are as follows: — 

“What is called jatavedcLJi (i.e., fire) the light which is 
within man, and the light called the Word which illuminates 
both the sentient and the insentient. In the last one is the 
whole of this world established, both that which moves and 
that which does not move.” 

What is meant by ‘the shortest route’ is: The Science of 
Grammar (laksana ), consisting of general and special rules , 2 
is the means of learning the whole of the word—Brahman in 
an easy manner, by means of the main rules and their elabo¬ 
rations. It is the means of inferring that those who, with¬ 
out being taught, are well acquainted with the true word 
endowed with correctness and free from corruption, are the 
cultured people . 3 Once it has been inferred who the cultured 
persons are, the Science of Grammar is the agent for declar¬ 
ing the correctness of words like prsodara on the ground that 
they are used by the cultured . 4 It has been said. 

“He who does not know the Science of Grammar does 
not know the word, the meaning, their mutual relation and 
the occasion for use nor what is correct and what is not, even 
when the meaning is the same nor who is cultured, inferred 
from their use of the correct forms of the words. ’ 

[1. It is the Word which appears as the objects. The objects are 
merged in the word, exist in the form of the word. Thus, there is 
the relation of cause and effect between words and things. The things 
existing within in the form of the word are the cause and the externa¬ 
lised word is the effect. Or, the inner word is the cause and the 
externalised objects are the effects. The same idea is set forth in 
the following verse, 



1. 13. 

2. Sdmunyavisesavaddhi laksanam . Vrsabha explains this as fol¬ 
lows — Utsargdpavdda - vidhipratisedlia - niptatanatidesasthanyddesa - 
linganiyamaih saptabhih sdrnanya-visesapralcdraili. As ten things are 
mentioned in the long compound word, but only seven things are 
meant to be conveyed, one can, according to the Ambakartri, under¬ 
stand the following: utsargdpavdda, vidhipratisedlia, vidliinipatana, 
vidhyatidesa , sthdnyadesa, linge samdnyavisesarupa and vidhiniyama . 

3. See M.Bha. on P. 6. 3. 109. where a description of Si§tah, cul¬ 
tured people, is given. 

4. P. 6. 3. 109.] 

13. The expression of what one wants to say (the 
principle of the use of things) depends upon words and 
the truth concerning words cannot be understood ex¬ 
cept through Grammar. 

(a) The basis for the expression of a meaning is that 
the speaker wants to communicate it, not whether it has an 
external existence as an object or not. Desire to communi¬ 
cate depends upon the existence of a suitable word. The 
speaker uses a separate suitable word for every meaning 
which his desire to communicate brings to the mind. It is 
like a person applying the right sense when he wants to cog¬ 
nise anything directly. 

(b) Another explains differently. What is meant by the 
truth ( tattva ) in the use ( pravrtti ) of an object ( artha ) is 
the ground for the application of a word to that object. When, 
due to the presence of that ground in an object, a cognition 
of that object corresponding to that ground arises, then one 
can bring that object into verbal usage. In the absence of 
that ground inhering in the object, there cannot be any verbal 
usage in regard to the bare object. As words acquire their 
form in relation to the universal (jdti) it is the universal 
(samdnya) which is the basis of words. 

(c) Or one can say that interconnection ( samsarga ) is 
the basis ( tcittvci ) of verbal usage. As the meanings of words 



are closely connected with one another, even though they 
appear to be unconnected, verbal usage depends upon the 
sentence. When the cognition of interconnection ceases, no 
verbal usage in regard to the word-meaning is possible. 

(d) Another meaning is this: A bare object comes only 
within the range of tyad (that) etc., which just point to 
things. The basis of its verbal usage is interconnection. That 
which is connected enters into relation with action in a pri¬ 
mary or secondary capacity. 

(e) Or activity ( pravrtti ) means an action like ‘being 
born’ conveyed by the verb. The essence (tattva) of this 
thing which is called activity is the fact of being something 
to be accomplished, its requirement of means, its assuming 
the form of sequence, its being the cause of suggesting time. 
The other object is a mere thing which remains the same 
in all the three times. It is conveyed by the noun and is 
free from all inner sequence. 

(f) Or again, what is the basis ( tattva ) of the verbal 
usage (pravrtti) of an object (artlia) ? The cognition, having 
the form of an object, which is projected as an external 
object. And that depends upon the word. 

What is meant by ‘the truth concerning words’ is its 
completeness (avaikalya) its correct form in which its purity 
has not gone. That is its complete form. The others, the cor¬ 
rupt forms, used by those who really intended to use the 
correct ones, are its incomplete forms. 

[The main word in this verse is arthapravrttitattvanam . This com¬ 
pound word is analysed in six ways in the Vrtti.: a. b, c, d; e; f. In 
a, it is analysed thus: arthasya pravrttitattvam ; in b, thus: arthasya 
pravrttau tattvam nimittam; in c, in the same way as in b; in d, thus: 
arthasya pravrttitattvam; in e, thus: arthasya pravrttesca tattvam; in f, 
thus: arthasya pravrttitattvam . In all the explanations except e, pravrtti 
means vyavahara = verbal usage. In e, it means action. 

The fact that as many as six alternative explanations have been 
given raises some questions which are discussed in the Introduction.] 


i. i4. 

14. It (Grammar) is the door to salvation, the re¬ 
medy for all the impurities of speech, the purifier of all 
the sciences and shines in every branch of knowledge. 

1 0ne, who, with a previous knowledge of the correct 
forms of words, realises the unity of the real word, goes be¬ 
yond sequence and attains union with it. By acquiring special 
merit through the use of the correct word, he is united with 
the great Word and attains freedom from the senses. 2 After 
having reached the undifferentiated state of the word, he 
comes to the source of all differentiation: Intuition (pratibha ). 
From that Intuition in which all Being is latent and which, 
due to the repetition of the union (mentioned above) tends 
to produce its result, he reaches the Supreme Source in 
which all differentiation in completely lost. What is meant 
by ‘the remedy for all impurities of speech’ is that it is like 
the science of Medicine (Ayurveda) in regard to the defects 
of the body. One who knows grammar does not use the 
corrupt forms which are the cause of sin. It has been said 

“Knowledge (of the correct forms) is his refuge” 
(M. Bh)a. I, p. 2, 1. 28.). 

What is meant by ‘purifier of all the sciences’ is that cor¬ 
rectness ( samskdra ) is known through it. In all the sciences, 
it is ths (correct) words which convey the meaning. It has 
been said: — 

“Whatever is learnt and not understood but is merely 
uttered as it is, is like dry fuel without fire; it will never 
burn.” 3 

There is also the following verse cited by those who 
have not gone astray: — 

“On the earth, water is the purest of things; of the 
(purified) waters, it is the sacred hymns ( mantras ) which 
are the purifying agents and the sages have declared that it 
is the science of Grammar which reveals the purity of the 
sacred hymns, Saman, Rk and Yajus.” 

22 vAkyAfadiyam of bhartrhari 

What is meant by ‘shines in every branch of knowledge* 
is that everybody follows the science of grammar even for 
composing a work on his own special subject and is very 
careful to avoid the use of corrupt forms. 

[1. This verse is understood in the Vrtti as describing the stages 
through which the grammarian goes in attaining moksa. Vr. says: 
etasyd eva brahmaprapteh kramam iddnlm varnayann aha avyavaklr - 
nam iti. Some, however, think that different forms of liberation are 
set forth here. 

One thing to be noted is that Bhartrhari, in the Vrtti, speaks 
about going from pratibha which is understood by Vr. as Pasyantl, to 
Para Prakrti. In other words, there is a stage beyond Pasyantl called 
Para Prakrtih . Pasyantl itself is called sarvavikdrdnam prakrtih. Thus 
Pasyantl is prakrti and after that comes para prakrtih. 

2. Vaikaranyam = apagatendriyatdm (Vr.). The same word is used 
in the Vrtti on verse 5 to describe one of the alternative conception of 

3. Ni. 1. 18.] 

15. Just as all the universals of things depend 
upon the form of their words for their communication, 
so is this science the basis of all the other sciences. 

[Sabdakrtinbandhandh. By akrti , it is the jati which is meant here. 
In other words, the universals of words convey the universals of 
meanings. (See Vak. III. Ja. 6). 

16. This is the first step in the ladder leading to 
liberation, this is the straight royal road for all those 
who desire salvation. 

17. Here the Self, being free from errrors and thus 
fit to study the Vedas, sees the very basis of the Vedas; 
their pure form. 

[ Chandasyah . This word can be explained according to P. 4. 4. 98: 
Tatra sadhuh 9 . It would then mean 'fit in regard to the Chandas, i.e. 

I. 18-22. 


fit to study it or protect it. Vr, however, explains it thus: Chandasyah 
iti. Chandasam samuhah , samuharthe aupasayikhyanikali taddhitali. He 
has probably in mind P. 4. 2. 42 according to which the suffix yan is 
added to words in the sense of collection.] 

18-22 4 That which 2 is the highest form of undiffer¬ 
entiated speech, the pure Light which appears differ¬ 
entiated only in the midst of all this darkness. 

The Light which those worship who have passed 
beyond the manifested state consisting of the cognition 
of things and actions and beyond light and darkness. 3 

That in which the manifestors of speech, like the 
signs of the alphabet appear, through Yoga preceded 
by the knowledge and use of the correct forms of words, 
like reflections in a mirror. 

That of which the different phonemes 4 of the 
Atharvans, the Samans, the Rks and the Yajus are the 

That which, though one, is variously interpreted in 
the different traditions, that supreme Brahman is 
attained by a knowledge of Grammar. 

[1. No Vrtti is available on verses 15-22. This naturally raises the 
problem whether they are integral parts of the Vakyapadiyam or quota¬ 
tions in the Vrtti on verse 14. Usually, quotations are immediately 
preceded by some word like tad yatha or evam hyaha which indicates 
that what follows is a quotation. There is no such word here. The 
fact that Vr. comments on them settles nothing as he comments on all 
obvious quotations also. There is, however, one indication that they 
are integral parts of the work: The word tatra occurring in verse 23 
refers to the science of Grammar mentioned in verse 22 which shows 
that the former is a continuation of the latter. As verses 18-22 form 
a group and constitute one sentence, verse 23 is a continuation, not 
only of verse 22, but of all of them. 

2. Some form of the relative pronoun yad occurs in all the verses 
from 18-22 and it is to be correlated \vith the word tad occurring in 



verse 22. What do these words refer to? to Brahman or to Vyakarana? 
Some of the points mentioned in the description seem to be equally 
applicable to both, i think, however, that they refer to Brahman. Vr. 
also makes it clear when he says in his commentary on verse 18. 

Brahma adhigamyate iti vaksyati. Tat sarvesiittaraslokesu sambandha- 

3. Vaikrtam murtivyaparadarsanavi samatikrdntah . Vr. explains: 
Yogina ityarthah. 

4. Pn'thaksthitiparigrahdli: Etad aha. ndndjdtiyavarnapratitibimba - 
parigrahena sthitah iti. (Vr.).] 

23. There, the great sages who are the authors of 
the sutras, varttikas and the Bhasya have declared 
words, meanings and their relation to one another to 
be eternal. 

The very basis of the science of Grammar is that the 
word, the meaning and their mutual relation are eternal. 
In it, by ‘word’ is meant the universal (dkrti) of the word 
It has been said: — 

“As the universal is eternal, so is the word eternal.” 1 
This science proceeds on the basis of the universal. In fact, 
it has been said: — 

That is already established, because it is the universal 
which is taught.” 2 

This universal is different from the particular universal call¬ 
ed wordness. Wordness is a universal which coexists in 
t e same thing with other (lesser) universals which cannot 
co-exist in the same thing. Universals of words, like the 
universal of the word tree, when there is vagueness, are 
mixed up with the causes of the manifestation of the indivi¬ 
dual word and when manifested by the individual word, are 
called words. 3 Just as, in a pot, the facts of being substance, 
earth and pot can co-inhere, in the same way, in the word 
tree, universals like those of being an attribute, a word 
and the word tree can co-inhere without mutual opposition, 

I. 23. 


(Objection) In the case of objects like the pot, the parts 
of which exist at the same time, one can see that the whole 
( avayavi ) is the cause of the manifestation of the particular 
universal (pot-ness). But the parts of a particular word 
cannot co-exist, they are not produced at the same time, they 
do not exist at the same time, they are unnameable ( avya - 
padesya) and so they cannot produce the whole word, the 
many material parts of which do not co-exist and where the 
universal can inhere. The universal word-ness exists in 
each part. If, similarly, it is maintained that the universal 
of the particular word tree also exists in each part, then the 
cognition having the form of the word tree would occur 
even after the first part is uttered, i.e., when parts like ‘v* 
are uttered in isolation (and that does not happen). 

This is not a valid objection. It is as in the case of 
actions like lifting, revolving, pouring etc., which are produ¬ 
ced and perish and whose parts do not produce another action 
corresponding to the whole (avayavi ). Nor is the inherence, 
m the parts, of universal like the fact of being lifting etc., 
distinct from the fact of being action, not accepted. 4 Nor do 
cognitions having the form of particular actions like lifting 
arise when only a part is seen. These parts of actions are 
due to special efforts and each of them is the substratum of 
universals like the fact of being lifting etc., but as this parti¬ 
cularity is difficult to grasp, they do not produce a cognition 
having a particular universal as its object. All the manifest¬ 
ing factors of that particular cognition have not yet come into 
being. Therefore, there cannot be any verbal usage based 
on it. But when these actions, involving contacts and separa¬ 
tions, determined by a particular direction, are perceived in 
succession, then verbal usage, characterised by a particular 
universal, becomes possible. In the same way, in the case of 
words like ‘vrksa’, distinct parts like V are produced by 
special effort, but their distinction is difficult to grasp and 
even though they, at the time of the utterance of each part, 
suggest the universals of particular words, one cannot perform 
any verbal usage with them, because universals do not, at 
that time, have many elements to suggest them. But when a 
K. S. 4 



succession of the parts is gradually perceived, verbal usage 
based on particular universals becomes possible. 

Grammarians do not necessarily accept the view current 
in other sastras regarding the mode of suggestion of the uni¬ 
versals . 5 Suggestors do not necessarily suggest what inheres 
in them. Even though the universal of a word may not inhere 
in the word, when the mind is prepared by the successive 
impressions left by the cognition of the previous phonemes of 
a word, the universal of the word, previously not cognised or 
indistinctly cognised, is perceived through the cognition of 
the last phoneme . 6 One infers the existence of the universal 
of a particular word from the recognition which one makes in 
the form ‘This is the same as that’ when words like c vrksa 9 
are uttered by parrots, sankas, men etc. 

Even those who do not accept the existence of such a 
universal declare that there is one eternal word which is 
suggested by the many sounds of a word . 7 Others still accept 
divisions in the form of phonemes within the word . 8 Some • 
others hold that the word is one, whether it be a phoneme, 
a word or a sentence, but appears to have parts produced in a 
sequence. Others still hold that, due to the continuity of tra¬ 
dition, there is constant usage and the speakers are not aware 
of the beginning of words which are eternal because of un¬ 
interrupted usage . 9 

The eternality of meanings is also accepted by some on 
the basis of the eternality of universals. It has been asked: — 

“According to what conception of word-meaning would 
the analysis “siddhe sabde arthe sambandhe” be proper? 
And the answer is: — 

“That it is the universal .’’ 10 

In this Bhasya passage, the eternity of meanings has been 
variously explained, according to all the views. It has to be 
understood according to the Bhasya. 

The relation is also eternal. What is meant in this: where 
there is the idea of mutual appurtenance, the relation between 
word and meaning in the form ‘It is this’ is. since meaning 


i. 25. 

cannot be assigned (by grammar) n , eternal, self-existing and 
not something not known before and made for the first time 
by some speaker for the benefit of some listener. Therefore 
the relation between word and meaning is beginningless and 
unbroken. Or, it may be stated that the relation between 
word and meaning is that of the illuminator and the 
illuminated (prakasyaprakasakabhava ), based upon con¬ 
vention ( samayopadhih ) a kind of fitness, like that 
between the senses and their objects. Or it may be 
said that the relation between word and meaning is 
causality based on unbroken tradition, considering that the 
cognition (arising from words) having the form of external 
objects, thought of as external and accepted as having a men¬ 
tal as well as an external object, just as the letters of the 
script are thought of as the phonemess of the alphabet, be¬ 
cause they bring the latter to the minds. It has been said 
(in order to show that the meaning is essentially mental): — 

“They, while explaining their deeds from birth to death, 
* make them present to the mind, objects of the mind.” 12 

What is meant by ‘taught there by the great sages’, is: 
by the authors of the sutras etc. Those who have composed 
the sutras etc. of the science of Grammar are referred to. The 
very fact that the sutras have been composed shows that 
they considered the words to be eternal. There would be 
no purpose in composing the science of Grammar if the words 
were not eternal. Because they would be a matter of mere 
usage and great cultured persons would not take the trouble 
of expounding them. Therefore, the science of Grammar pro¬ 
ceeds only when the correct forms of words are well estab¬ 
lished. Others bring forward sutras like 

Tadasisyam samj iiapramanatvcit . 

“This, (the concord of gender and number taught in 
P. 1, 2, 51.) need not be taught because names are to be 
accepted as they are (praiivana) .” 13 

as proof that the words are eternal. In the varttikas also, 
there are the following statements which show the eternality 
of the word. 



“On the basis that the word, the meaning and their rela¬ 
tion are eternal.” 14 

“That is already proved because the word is eternal.” 15 

“It is sphota which is the word, the sound is only the pro¬ 
duct of effort” 16 ( vyciycima ). 

“The whole word takes the place of the whole word.” 17 

In the Bhasya also, it has, indeed, been said: — 

“This matter has been specifically considered in the 
Sangraha, namely whether the word is eternal.” 18 

It has also been said in a Bhasya passage. 

“Word being eternal, the phonemes also should be eternal 
and changeless.” 19 

Even if eternality is taken as something merely practical 
(the teaching of augments and substitutes does not violate it). 
As has been said: — 

“(As) both Khadira and Barbura have fine leaves and 
yellow stalks, (when one says) ‘the Khadira has thorns’, 
(what happens is not that the statement adds thorns which 
were not there, but that the idea of Khadira which might 
have extended to both is now restricted to Khadira only.) 20 
Similarly, when one, after having said: — 

“To the east of the village are the mango-trees” (one 
adds), “the banyan trees have milk, downward growths and 
wide leaves”. 

(all that happens is that the idea of mango-trees which 
might have been wrongly extended to the banyan trees 
also now disappears and, in its place, the idea of banyan 
tree comes). 

Or (it might be said) those very sages who have realised 
the truth and have, in the course of their different teachings, 
composed Sutras, Anutantra (varttikas) and Bhasya, have, 
in the science of Grammar also, declared that the word, 
the meaning and their mutual relation are eternal. And their 
authority in the world is established. 

[1. Though the idea is found expressed in the M.Bha, the actual 
words are not traceable. 


I. 23. 

2. M. Bha I, p. 13. 

3. gabdakrtivisesa hi etc. This sentence is somewhat obscure. 
The previous sentences told us that the word stands for the universal 
and that the universal of a word can co-inhere in it with other lesser 
universals. If the word stands for the universal, why do we say ‘word’ 
instead of saying ‘wordness’? This sentence is apparently an answer 
to that question. The main part of the sentence is vrksasabdatvadayah 
sabdd ityapadisyante = The universals of words like Vrksasabdatva are 
merely referred to as the words. That is because they are identified 
(sarupatam apanndh ) with the cause of their manifestation ( nimitta), 
namely, the individual words. This manifestation is necessary. Other¬ 
wise there would be extreme vagueness (sati vastusampramohe) and 
there would be no verbal usage at all. 

4. Karmatvasamanyadanye utksepaiuitvddayaste§u karmaksane§u 
samaveta abhyupagamyanta eva vaisesikaili —Vr. 

5. Na cavasyayi sastrantare parifosta jatyabhivyaktiprakriya vaiya- 
karanaih parigrhyate. The usual view connected particularly with 
the Vaisesikas, is that the universal (jati) inheres in the particular 
which, therefore, reveals it. Though the universal is eternal according 
to them, it exists in the particular and it can be perceived only as 
existing in the particular. The latter, therefore, is said to reveal it 
(abhivyahjaka ). When such a process is described, universals like 
ghatatva , existing in ghata, are kept in mind. The universal vrksa - 
sabdatva does not exist in the word vrksa exactly as ghatatva exists in 
the ghata and its perception is also a different process which is des¬ 
cribed in greater detail in the course of the work. For the present, 
it may be noted that when light and the senses reveal objects, they 
reveal things which do not inhere in them, but are external to them. 

6. There are two views on this question: (1) cognition of the 
last phoneme, accompanied by the impressions left by the cognitions 
of the previous phonemes, cause the cognition of the universal, (2) the 
cognition of the last phoneme, together with those of the previous 
phonemes, leave impressions which cause the cognition of the universal. 


7. Yair apyakrtivyavaharo nabhyupagamyate etc. The same word, 
uttered by different persons, at different times or even by parrots etc, 
is recognised as the same word. But this recognition of identity 
must not be confused with the universal. If it were the universal it 
should rest on a particular word as ghatatva rests on. a particular 
ghata. But it seems to be somewhat independent of it. So it is con¬ 
ceived by some as an eternal independent entity and not as a universal. 
The fact that it is suggested by the impressions left by the different 
phonemes of a word also points to the same conclusion. 

8. Vr. remarks: parasparavyavrtta vary&statreti bhdgav ant am 
sphotam manyante . 



9. Vyavaharaniiyatayd nitytih sabdd Hi. Vyavaharanityata, vyavas- 
thdnityatd,, all mean continuity as distinguished from 
K-utasthanityatd, absolute etemaliity or changelessness. 

10. M.Bha. I, p. 7, 1. 8. 

11. Arthanadesanat. tacca laghvartham. ko hi samartho dhdtupmti- 
padikapratyayanipdtdndm arthan ddestum. (M.Bha. I, p. 363). 

12. M.Bha. on P. 3.1.26. 

13. P. 1.2.53. 

14. M.Bha. 1.1. 

15. Not traceable. 

16. Not found in the Varttikas nor in the M.Bha. 

17. M.Bha on P. 1.1.20 (I, p. 75). 

18. M.Bha. I. p. 6, 1. 12. 

19. M.Bha. I. p. 18. 1. 14-15. 

20. M.Bha. I. p. 113. 1. 12.] 

24.-26. The meanings which have been obtained 
by abstraction, those which are of a fixed character, the 
forms which have to be grammatically analysed and 
those which are used as means for this purpose, the 
relations consisting in causality and fitness which lead 
to merit and understanding of meaning in the case of 
correct forms and to the understanding of meaning only 
in the case of incorrect ones, these have been described 
in this sastra through indications and direct statements. 
These are only some which have been dealt with here 
according to tradition, in order that they may be fixed 
in memory. 

These three verses give the whole of the subject-matter 
to be dealt with. The abstracted word-meaning is tha't which 
was closely connected, but being freed from that connection 
by an inferred postulated form, it is now abstracted. The 
form of that separated thing is beyond the range of usage. 
It is generally established on the basis of postulation, by 

I. 24-26. 


following tradition, ' according to one’s understanding and 
through repeated practice. Similarly, the essence of the word 
being indivisible, in order that the work of Grammar may 
proceed, by adopting the method of positive and negative 
reasoning and postulating their recurrence, the words are 
separated from the whole and the separated word-meaning 
is looked upon as what is expressed by them. This sepa¬ 
rated word-meaning serves the usage of the science of Gram¬ 
mar and also worldly usage based on difference and is simi¬ 
lar to usage in Grammar. When this separated meaning rests 
on one single word, it cannot be determined as true or false. 
When the mere word tree ( vrksa ) or fig-tree ( plaksa ) is 
uttered and it is not completed by the addition of a verb, 
the meanings conveyed by the words cannot be clearly de¬ 
fined. As long as specific actions which keep out others do 
not set aside the act of mere existence which is the cause 
of the very use of a word, 1 the verb ‘it is’ ( asti ), in the third 
person, denoting existence, though not used, is brought to 
the mind by the words tree etc., and is understood. These 
words, looking like single words, when completed by some 
word or other, are called sentences. Similarly, such divisions 
as the meaning of the first word (of a compound) the mean¬ 
ing of the second word, the meaning of an outside word, the 
meaning of the stem, the meaning of the root, the meaning 
of the suffix etc., are abstracted somehow in many ways from 
a single word by some scholars without any clear line of 
demarcation. 2 The meaning with a fixed character is that 
which is conveyed by the sentence, it has fictitious divisions^ 
it is specific, one, of the nature of action. It is conveyed by 
means of the cognitions of the separated word-meanings. 
Even though the separated elements are cognised, the mean¬ 
ings understood from namasyati , (he pays homage) scingra- 
mayate (he offers battle) mundayati (he shaves the head) 
kuttayati (he powders) carvayati (he masticates) etc., at 
the time of the comprehension of the whole, is not complex. 3 
That is why it has been said: — 

“Or it is not necessary, because it is the (complete) word 
which is used to convey the meaning.” 4 



Word’ here stands for that in which an action is merged. 
Or one might say that it is only the apparent persistence of 
the cognition ( samjpratyaya ) (of the separated elements) 
which is referred to here. Accepting the use of the indivisi¬ 
ble) word to convey the (indivisible meaning, the Bhasya- 
kara) again says: — 

“It is the word which ends in a primary or secondary 
suffix which has a meaning and not the bare primary and 
secondary suffixes.” 5 

What is meant by ‘forms to be grammatically analysed’ is 
this: According to some, it is the individual word which is the 
limit of grammatical analysis, while according to others, it is 
the sentence which is the limit. 6 According to the view that it 
is the word which is so, when, because of the identity of 
sounds, the same form of the word is taken, words which 
have acquired their correct form on the basis of the univer¬ 
sal, even when the particulars come on the scene because 
of connection with other words, would continue to have the 
form based on the universal, which is an inner factor. The 
result would be that a word like isuklci (white), in the sin¬ 
gular number and neuter gender, would be connected with 
other words expressive of the substratum, having some other 
gender and number. In order that this may not happen, the 

“Of the adjectives also, except the universal”. 7 

lays down that when words expressive of the substrata 
which are external factors are to follow, words expressive of 
quality, should take the gender and number of the words 
expressive of the substrata. According to the view that gram¬ 
matical analysis has the sentence as the limit, considering that 
a quality always exists in something, it is not possible to sepa¬ 
rate it from its substratum and so, being fully determined, 
it does not denote the general idea at all. On this point, it 
has been said: — 

“That is natural.” 8 

I. 24-26. 


Similarly, it is according to the view that each word in a 
‘dvandva’ compound gets its correct form separately that the 

“In a dvandva compound, the gotra suffix must be elided 
even in the numbers other than plural?” 9 

has been taught. On the view that it is the whole which 
gets the correct form, it has been said: — 

“Or it may not be taught, as, in a dvandva compound, 
every word is in the plural number.” 10 

Similarly, all the sutras beginning with 

“A qualifying word ending in a case-affix, is variously 
combined with a qualified word ending in a case-affix” 11 and 

“Words expressive of the standard of comparison and 
ending in a case-suffix are combined with words expressive 
of the common property ending in a case-affix,” 12 

are to be understood as having been composed on the view 
that the word is the limit of grammatical analysis. It is after 
accepting that it is the word which has to be grammatically 
analysed that such divisions as bhu ti y bhu ati , leading to the 
understanding of groups of other words (than the sentence) 
have been postulated and accepted as means ( pratipadaka ). 

What is meant by ‘causality’ is this: when the cognition 
which has parts of the appearances of the object and is super¬ 
imposed on the object is taken as the object, then, of that 
object, the word is the cause. Since the relation of identity 
between the word and the meaning in the form ‘It is this’ is 
established, that cognition of the meaning is the cause of the 
application of the word which is within and which is mani¬ 
fested by the sounds. Because of the relation of illuminator 
and illuminated between particular words and particular 
meanings, as in the case of the senses and their objects, there 
is an eternal fitness, not created by anybody, of the express¬ 
ive words possessing unvarying and well-established correct¬ 
ness in regard to the meanings expressed by them. 13 In 
the case of words whose relation with their meanings is not 
established at the time of their first application, the fitness 
depends upon convention. 

K. S.5 



The relation of the correct word with its meaning be¬ 
comes auxiliary to the understanding of the meaning and 
when it is used with a knowledge of Grammar, 14 it becomes 
auxiliary to the manifestation of merit. In the production of 
special knowledge of the meaning, it brings about a condition 
similar to perception. Corrupt forms, on the other hand, be¬ 
come, like winking etc., auxiliary to the production of special 
knowledge in the manner of inference, by their connection 
with what is connected with the meaning. 

What is meant by “These have been described through 
indications” is this:—When, in order to carry on the work 
of Grammar, the analysis of some unified meaning is made, 
there can be difference of opinion among men regarding the 
different limits of the (analysed) meanings. How is it to be 
known whether the suffix ‘nic’ is to be added to the root 
when the causative is to be expressed or whether it is to be 
added to the root of which it is the meaning? 15 Similarly, 
in regard to the sutra 

“When the idea of the feminine is to be expressed” 16 
the doubt arises whether the suffixes ‘tap’ etc., are to be 
added to a nominal stem when the idea of the feminine 
gender is to be expressed or to a stem which includes the 
feminine gender in its own meaning. Similarly, the doubt 
arises: the meaning of which element is important in a com¬ 
pound-word formed according to the rule. 

“The negative particle c na’ may be compounded with a 
word connected with it in meaning and the resulting com¬ 
pound word is to be called tatpurusa.” 17 

These options do not exist in the world, because the 
worldly meaning is the whole and in regard to that, there is 
no deviation. The alternatives adopted by men being thus 
indefinite, that option is accepted by which the decisions of 
the Science of Grammar are not affected. Similarly, that 
the meanings of the inflectional suffixes are the numbers 
one etc., or the accessories to the actions like the object, that 
the meaning of a nominal stem is the group of five or of 
four or of three, such alternative views are due to the varia¬ 
tions in human intelligence. 18 Action, accessary to action 

i. 24-26. 3S 

and time are also analysed as the expressed meanings in 
different ways by different people. It has been said: 

“The root stands for the accessories, action and time, 
person and number, that is the verb; gender and number, 
accessories to action, this is what the nominal stem stands 
for.” 19 

That the tatmanepada suffix comes after a root which 
expresses mere action or the object of action, that the paras- 
maipada suffix is added to a root which expresses the agent 
has been said on the basis of an imaginary separation. The 

“After conveying its own meaning (the universal) which 
inheres (in its substratum) the word, (though) free from any 
further requirement, denotes the substance” 

is nothing more than the following of the order in which 
the understanding takes place. 20 A word does not convey its 
own meaning and others by pausing again and again, because 
it is uttered only once. Nor is it ever separated from its mean¬ 
ing. Nor is there any fixed sequence in the understanding of 
the meaning by the hearer or the speaker. 21 The object, 
qualified by all the attributes, a bundle of all the parts . which 
are closely linked, comes at one and the same time- within 
the range of a single cognition and later, one deliberately 
analyses it into different cognitions. But, since an intuition 
leading to purposeful activity cannot take place without the 
reunification of what has,, been divided, one again under¬ 
stands the connected form. The sequence of the cognitions 
of the speaker or the hearer who understands the parts sepa¬ 
rately, through considerations of proximity, width of scope, 
abundance of the causes of manifestation, the desire to know 
and the tendency to awaken the seed of another cognition 22 
is not fixed in regard to the parts which are to be understood. 
For it has been said: — 

“The one object is divided in many ways on the basis 
of difference of powers, by following the different forms of 
cognition, by those who know the nature of knowledge.” 



The meaning with a fixed character in the Science of 
Grammar is either the word-meaning or the sentence-mean¬ 
ing. For it has been said: — 

‘Or no, for it is the word which is used to convey the 
meaning.” 23 and 

“What is extra, that is the meaning of the sentence.” 24 
In the Sangraha also, it has been said: — 

“What is called the word is not fixed in its form. The 
form and the meaning of a word are born from the mean¬ 
ing of a sentence.” 

Even though the form of the word to be explained is 
determined, according to the view that the sentence is the 
limit of analysis, there is no fixity in the form to be accepted 
when it comes to division into stem and suffix, of words like 
viarutta , indrarii, aikagarika, girisa 3 srotriya , ksattriya etc. 25 
It has been said: — 

The word with its meaning comes from the meaning 
(of the sentence). It is through the word that the meaning 
of a sentence is determined. The sentence is born of a collce- 
tion of words and the word is born of a collection of pho¬ 

At the time of derivation, it is stated that the word 
exists in the sentence and that the word is born of a collec¬ 
tion of phonemes. 

The relation has also been shown as that of cause and 
effect. For example: — 

‘They, describing their fortunes from birth to death, 
reveal them as existing, as figuring in the mind.” 26 

Here and there, the relation has also been described as 
fitness. For example: — 

Expression, on the other hand, is natural.” 27 While the 
colour remains the same, the words sona, karka and hema 
are applied to a horse and not to a cow and others. 28 

Similarly, it has been said: — 

t 24-26. 


“Among those who work and study equally, some are 
rewarded with understanding while others are not” 29 
and so on. 

“Such roots having the circumflex accent and n as indi¬ 
catory letter have been taught by the Teacher which are 
both, i.e., the fruit of whose action is meant for the agent 
and not meant for the agent.” 30 

The author of the Sangraha says: — 

When the word and the meaning are separated, there 
is division in usage. Because (really speaking) the unity of 
the word and the meaning is established.” 

Again he says: — 

Neither in the world nor in the Veda is there any per¬ 
son who creates the relation. How can the relation of words 
(with meanings) be created through words?” 

[1. Yeyam aupacariki satta sa sarvasabdapravrttikdranam (Vr.). 

2. It is not merely the indivisible meaning of the sentence which 
is artificially divided but the meaning of a single word can also be 
so divided. The addition of the verb asti to the single word vfksa 

s ows that even sentences consisting of single words are artificially 

3. Namasyati etc. Even though namasyati is explained as namah 
karoti, ‘he pays homage’ as involving an agent’ an object and an 
action, what is understood from it in the end does not involve such 

fferentiation, but is a unified whole. The separation of the elements 
at the time of the explanation is artificial. When grammar speaks about 
primary and secondary suffixes and their meanings, it is all artificial. 
It is the whole word ending in one of these suffixes which has a 
meaning. See P. 3.1.17, 19, 21; 25. 

4. Va. 19 on P. 1.2.64. (M.Bha. I. p. 237). 

5. M.Bha. I. p. 319, 1. 6. 

6. Padavadhikam anvakliydnam. The word sukla ordinarily de¬ 
notes the quality white. When one wants the word to denote, not 
the quality, but the thing which has that quality, the suffix matup 
has to be added to the word according to P. 5.2.94. But the elision 
of this suffix after words expressive of quality has also been taught. 

, Us the word means ‘a white thing’ the form would still be 

sukla. Thus the word sukla denotes two different things: the quality 
white and the thing having that quality. The fact of being white 



(suklatva ) exists in both and, on that basis, the word gets its form; 
namely, its singular number and neuter gender: suklam. Even when a 
word expressive of the thing or things in which the quality resides 
is used like patdh , the word suklam which has already acquired its lorm 
according to the view that the word is the limit of grammatical analysis, 
would retain it and we would get the expression: suklam patdh. But as 
the expression is wrong, Panini, in order to prevent its occurrence; says: 
Visesananam cajateh (P. 1.2.52). According to this rule, adjectives take 
the gender and number of the words which they qualify. The very fact 
that he makes such a sutra shows that he considers the word to be the 
limit of grammatical analysis. 

Vdkyavadhikam anvakhydnam. According to this view, a quality 
always resides in its substratum. It cannot be thought of in isolation. 
The gender and number of the word expressive of the substratum would 
inevitably belong to the word expressive of the quality. Therefore 
words like sukla never express the quality white, in isolation. So it 
would never have the singular number and neuter gender in isolation. 
It would have the gender and number of the word expressive of the 
substratum where it resides. Therefore, no special sutra is neces¬ 
sary to bring it about. That is why it has been declared unneces¬ 
sary in the sutra Tad asisyam sampwipramdnatvdt (P. 1.2.53). 

7. P. 1.2.52. 

8. M.Bha. I. p. 430, 1. 11. 

9. Va. 5. on P. 2.4.62. (M.Bha. I. p. 490). 

10. Va. 8. on P. 2.4.62. (M.Bha. I. p. 491). 

11. P. 2.1.57. 

12. P. 2.1.55. 

13. See the Vrtti on Ka 23. 

14. Sabdapurvake va prayoge. See Va. 9, Paspasahnika and the 
M. Bha thereon. But sabdapurvaka prayoga which leads to dharma and 
abhyudaya must be distinguished from sabdapurva yoga which is men¬ 
tioned in several places in the Vrtti. See the Vrtti on Ka 14, 131, 142. 
It must also be distinguished from Vdgyoga which is already mentioned 
in the M. Bha, Paspasahnika and here in the Vrtti on I. 130. 

15. P. 3. 1. 26. 

16. P. 4.1.3. 

17. P. 2.2.6. 

18. There are different views current among grammarians as to 
the number of meanings which a bare stem ( pratipadika ) can convey. 
Some hold that it can denote five things ( pncaka ): svartha (the uni¬ 
versal), dravya (the particular) linga (gender), samkhyd (number) and 
accesory to action ( kdraka ). For example, in the sentence dadhy 
dnaya, the word dadhi denotes all the five things, even though, in form, 

I. 24-26. 


it is no more than the bare stem. Others think that the accessory to 
action is conveyed by the suffix and so attribute only four ( catuska ) 
to the bare stem. Others think that number also is the meaning of 
the suffix, in which case only three ( trika ) remain for the bare stem. 
If the feminine gender is considered to be the meaning of the feminine 
suffix, the bare stem would convey only two ideas ( dvika ). 

19. According to Vrsabha, this is a sutra belonging to the Kasakrtsna 
school. He is also our sole authority for the meanings of the technical 
words used in the sutra. From his explanation it follows that, ac¬ 
cording to Kasakrtsna, the suffixes do not express any meaning apart 
from what the root or the nominal stem expresses. As Vr. puts it: 
pratyaydndmavyatiriktdrthdbhidhdndt. It is also to be noted that, in 
the supposed sutra of Kasakrtsna, which verbal element expresses 
which meaning is directly stated, so that one does not have to depend 
on indications. 

20. Pratipattikramaniyaindnugamamdtram—Na sabdagato’ abliidhd- 
nakramo’ pi tu pratipattyapratipattikramah (Vr.). 

21. Pratipattikramo liy ayam srotur abhidhdtur vd na vyavastliitah. 
The idea is that there is no sequence at all. The following sentence 
makes it clear. 

22. In understanding the parts abstracted from the unified whole, 
the speaker or the hearer is guided by the following factors, as explain¬ 
ed by Vrsabha:—1. Pratyasatti= proximity. It is the universal which 
exists in a particular that helps one to distinguish the latter from other 
kinds of particulars. So that universal is proximate to the particular 
and, therefore, it is understood first. But the universal cannot be 
understood except through the particular in which it inheres nor can 
gender be understood except through the particular in which it resides. 
So, through the desire to cognise the universal and proximity to gender, 
the particular is grasped. Gender is understood before number and 
accessory because it does not depend upon any other particular for its 
comprehension and it also helps to distinguish the particular in which 
it exists from other particulars; 2. Width of scope (mahavisayatva) 
can be seen in the universal which inheres in all the particulars. 
Compared to gender, the particular has a wider scope, because all the 
genders can exist in the particular. One gender excludes the others 
and so it has a narrower scope. Gender is wider in scope than number, 
because the same gender can cover all numbers, whereas one number 
excludes the other numbers. As all the particulars can manifest the 
universal, the latter has abundant causes of manifestation; 3. (ablii- 
vyaktinimittopavyanjanaprakarsa) . The universal has more mani¬ 
festos compared to the particular which i s manifested by its 
own parts only; 4. The desire to know ( upalipsa ) is also a 
factor: what is desired to be known first is known first. Finally, 
all cognitions have inner dispositions ( blja ) among their causes. They 
are awakened ( vrttildbha ) before the cognitions take place. The cog- 



nition of the universal tends to awaken the predisposition to cognise 
the particular while that of the latter tends to awaken the predisposition 
to cognise gender etc. This is what is called; 5. bljavrttildbhdnugunyam . 

23. Va. 19, on P. 1.2.64. (M.Bha. I. p. 237). 

24. M.Bha on Va. 2. on P. 2.3.49. (M.Bha. I. p. 462). 

25. These words can be analysed in different ways — 

Marutta: Marut + tap (Va. 10 on P. 5.2.122) or Marut + dd -|- kta, 
in the sense of marudbhir dattah, the root dd being elided. 

Indrdnl: Indra + dnuk -J- nls or Indra + an (causative of an) + 
an -f- nip. 

Aikdgarika: Ekagdra -f thant or Ekagara + than. 

Girisa: Giri + si -f- da ( aunddika ) or Giri -f- sa (matvarthiya). 

Srotriya : nipata according to P. 5.2.84. or Chandas srotra + 

ghan (adhlte ityarthe). 

Ksatriya: Ksatra -f- glia or Ksatra + iya. 

26. M. Bha on Va. 15 on P. 3.1.26. 

27. Va. 15 on P. 2.2.29. (Vol. I. p. 433). 

28. Cf. M.Bha. Vol. I. p. 433. It is there pointed out that the words 
sona, hema and karka which mean red. black and white respectively, 
can be applied only to a horse of that colour and not to other animals 
of the same colour. 

29. M.Bha. on P. 2.2.29. (Vol. I. p. 430, 1. 12). 

30. M.Bha. on P. 1.3.72. (Vol. I. p. 292, 1. 22). 

27. The correct words, acquired from the cultur¬ 
ed through tradition, are the means of obtaining merit. 
The incorrect words, while not differing from them as 
far as conveying of the meaning is concerned, are of an 
opposite character. 

Just as other means of attaining merit, received through 
uninterrupted tradition consisting of successive teaching of 
the cultured, or just as prohibited things like injury to others, 
telling lies, stealing etc., or things neither allowed nor pro¬ 
hibited like hicup, breathing, scratching etc., are well estab¬ 
lished and are not to be doubted, in the same way, the dis¬ 
tinction between what is correct and what is not is based on 
unbroken tradition and is not to be doubted, but is to be 
accepted according to tradition. 

I. 28. 


28. Whether words be eternal or otherwise, their 
beginning is not known. As in the case of living be¬ 
ings, there is what is called continuity ( vyavastha- 

1 Whether one accepts the view that the words are eter¬ 
nal and manifested or the view that they exist beforehand 
and undergo modifications like birth etc., or the view that 
they do not exist beforehand but are made audible and be¬ 
come inaudible again (sopakhyanirupdkhyatvani ), in any 
case, there was no first stage in which verbal usage did 
not exist nor will there be a future stage when it will not 
exist. For those who declare that God, Time, the Purusas, 2 
knowledge, the individual soul ( ksetrajna ) to be different 
(from the effects and their causes), for those who declare 
this (universe) to be without God, Time, the Purusas, know¬ 
ledge or the individual soul but entirely based on nescience, 
for those who maintain that the universe, without a before 
and after and without going beyond the unity of the ultimate, 
consists of the manifestation of the different forms due to the 
combined, mutually opposed and indefinable powers, for all 
of them, there was no first period of time when there was 
no activity of the living. This is what is called continuity, 
without a beginning and without an end. It has been said— 

“That is also eternal, in which the identity is not des¬ 
troyed.” 3 

[1. According to Vrsabha, two views are here expressed in regard 
to words: (1) that they are eternal and, therefore, already existent °and 
manifested at the time of use, (2) that they are produced at the 
time of use and disappear again. The first view is associated with the 
Sarikhyas, especially Varsaganya. The second view seems to be that 
of the Naiyayikas. In regard to the universe, the view that it is with¬ 
out God etc., but based entirely on nescience is attributed by Vr to 
Brahmavadinah. Some, however, attribute it to the Buddhists and the 
view expressed in the latter part of the sentence to the Advaitins. 

2. Purusa — Purusa iti kecit prthivyadisamghatam ahull (Vr.). 

3. M.Bha. I. p. 7, 1. 22.] 




29. Nobody would establish this system of rules 
without a purpose. Therefore, this tradition relating to 
correctness is being composed by the cultured. 

Which cultured person, even if he has a confused mind 
( sambhinnabaddhih ) and no sympathy towards the world, 1 
would undertake the regulation of the accent and othei 
signs of correctness of the words of the Veda and of the 
world which are so difficult to know and to learn and which 
are the very purpose of Grammar? And such a legulation 
would not be useless. (If it were) such a regulation made 
by the cultured would be unacceptable to the others. And 
it would not be authority in the world for scholars. There¬ 
fore, this tradition (relating to words) beginningless, hand¬ 
ed down from teacher to pupil, the means of inferring who 
the cultured persons are, infallible, consisting of general lules 
and their elaborations, is being composed in different ways, 
through direct statements and by implication. 

[1. Sambhinnabuddhir api lokam pratyaYiabhinivistali, The reason¬ 
ing behind this expression is not clear. The purpose of this verse is to 
justify the composition of the science of grammar. Four reasons are put 
forward justifying its composition: (1) that it is eternal, (2) that it 
has been handed down from teacher to pupil, (3) that it helps us 
to infer who is cultured and who is not, (4) that it gives an infallible 
knowledge of correctness. Those who are confused as to what is right 
and what is wrong ( sambhinnabuddhih ) and have no sympathy for the 
world ( lokam pratyanabhinivistah ) would not be able to undertake 
the composition of the science of grammar. Vr. seems to have had 
the reading: lokam pratyabhinivistah which can mean, ^ according to 
him, “ api namayam lokah kliedaydsabhyam yajyeta iti vyutthita - 
buddhir abhinivistah = one who would like the world to be troubled 
by pain and fatigue. But Vr. considers the former reading easier to 

30. Dharma cannot be determined by reasoning 
alone, without the help of tradition. Even the know¬ 
ledge of the Seers is due to their previous observance 
of the tradition. 


I. 30-3i. 

All thinkers, when they reach the extreme, have recourse 
to the own nature of things. The determination of the own 
nature of actions having invisible fruits can be made only 
with the help of tradition. How can one trust human reason¬ 
ing in which the similarities and differences of things are 
never certain and which is, therefore, always doubtful? Even 
in the case of those individuals who have adopted a parti¬ 
cular mode of intellectual and spiritual life and are known 
to have acquired qualities which reasoning cannot explain, 
it is said that, because of their adherence to the injunctions 
of the Agama (Tradition), their souls are purified and a 
divine knowledge is manifested in them. 1 To assume that 
such knowledge is natural to them would be to condemn all 
special effort 2 as fruitless and obstacles to such knowledge 
would also arise naturally. 

[1. Sage Kapila is usually given as the example of one, who posse¬ 
ssed divine knowledge capable of seeing the past, the present and the 
future, the subtle ( suksina ), the distant (viprakrsta) and the hidden 
(vyavahita ). 

2. The effort made by persons to acquire the kind of knowledge 
that Kapila had is meant here. Vr. says: yatna ity abhyasadvkali.'] 

31. Nobody can violate, on the basis of reasoning, 
those paths of Dharma which have come down without 
a break, because they are accepted in the world. 

In spite of many variations in the doctrines of the cul¬ 
tured, there are well-known beneficial modes of conduct, 
common to all; to go against them would not be liked by the 
people. They have never been refuted by mere reasoning. 
It may be that some do, on the basis of some passage in the 
tradition itself, 1 resort to a reprehensible mode of conduct, 
opposed to what is accepted in the world. 

[ 1 . Te’pi hi vedantadarsanam anyatha parikalpya pmvrttali—Vv. Ac¬ 
cording to this, it is some passage of the Vedanta, i.e., the Upanisads, 
which is misunderstood by some people and made the basis of 
reprehensible conduct. Could it be such a passage as na kahcana 
pariharet (Cha. Up. 2.13.)] 


vakyApadiyam of bhartrhari 

32. It is extremely difficult to establish by reason¬ 
ing the nature of objects, because their properties differ 
according to difference in circumstances, place and 

The inference, after seeing what is connected and consi¬ 
dered invariably concommittant or something similar to it, 
of another thing which is not visible, does not lead to any 
certainty in regard to that invisible thing. One sees that 
the nature of things of proved strength and quality in cer¬ 
tain circumstances, varies, in other circumstances, visible or 
invisible to ordinary persons. Even the properties of exter¬ 
nal things like seeds and plants are suspended in certain cir¬ 
cumstances. Similarly, properties of things change with 
change of place. Waters of the Himalayas are very cold to 
the touch. But similar waters found in clouds, a water- 
boiler and the like are warm to the touch. The ordinary man 
( arvagdarsana ), misled by external resemblance, is unable 
to see the difference and can see it only with the help of 
tradition. Similarly, properties of things change with time. 
The temperature of the waters of a well and the like, is very 
different in summer and winter. Which intelligent man 
would ’ try to demonstrate, by mere reasoning, this subtle 
difference in nature, imperceptible to the ordinary man, un- 
ascertainable by inference and incomprehensible except 
through knowledge derived from tradition? 

33. The known power of an object to produce 
different effects is inoperative when it comes into contact 
with particular objects or factors. 

The proved power of fire etc., to bring about destruc¬ 
tion in wood etc., is inoperative in regard to such things as 
masses of clouds. Similarly, due to the action of magical 
formulae, special herbs or chemical products, the power (of 
fire) to burn even inflammable objects is suspended. Subs¬ 
tances whose powers are proved in some cases are of doubt¬ 
ful effectiveness in other cases. 


1. 34. 

34. Whatever is inferred with great effort by 
clever reasoners is explained otherwise by cleverer 
ones. 1 

Substance is different from quality because the latter 
qualifies the former. It is like this. Even though there would 
be the difference of qualifier and qualified in both cases, the 
word ‘king’ can qualify the word ‘kingdom’ (as in the phrase 
‘the king’s kingdom’ = rdjno rastram) but not the word 
‘hermit’. The word sandal can qualify the word ‘smell’, but 
not the word colour. Therefore, substance is different from 
quality. Even though the separate identity of substance is 
thus established, it has been said that the argument is not 
sound. The fact is that words sometimes convey the general 
and sometimes the particular. To qualify the word smell or 
the like with the word sandal which stands for something 
having a particular colour, shape etc., has a purpose. To qua¬ 
lify it with the mere word colour would be useless, because 
colour is a general word and as much as it conveys was alrea¬ 
dy known. When the question ‘whose man is this’ is asked 
the answer ‘the king’s’ is given in order to exclude other 
possible masters. Nobody would answer ‘of some person’, 
because that much was already known. Others give another 
example: Even though a word and a stanza have this much 
in common that both are parts of a hymn (which is a collec¬ 
tion of words), a stanza can qualify a word (as in the ex¬ 
pression ‘the word of a stanza’ = rcah padam) but not vice- 
versa (one would not say: the stanza of the word). Another 
argues differently: It is on the assumption of identity that 
the word sandal qualifies the word smell and the word colour 
does not. So the reason given: ‘because the latter qualifies 
the former’ to prove difference (between substance and 
quality) should prove the opposite. The problem arises be¬ 
cause of particular meanings which figure or do not figure 
in the mind when words are uttered. Therefore, the infe¬ 
rence of something which is invisible from what is visible 
is alright if it is not opposed to tradition. It is a fact that 
minor acts, closely connected with particular major acts like 



cooking and ordinarily leading to their inference, may be 
performed by some for deceiving others. 2 

[1. Cf. Tatha hi kaiscid abhiyuktair yatnenotpreksitastarka ablii - 
yuktantarairanyairabhasyamand drsyante —Sank. Bha. on Br. Su II. 

1 . 11 . 

2. The purpose of this karika is to show that reasoning, unsupport¬ 
ed by tradition, is unreliable and that the reasoning of one may be 
upset by another. The argument of the Naiyayika to prove tliat 
substance is different from quality is taken as an example. His argu¬ 
ment, briefly, is: Substance is different from quality because the latter 
can qualify the former. Against this, others point out that the argu¬ 
ment is weak because both what is different and what is not different 
from another can qualify it in some cases and cannot qualify it in other 
cases. A king and a kingdom are different from each other and we 
can say: the king’s kingdom (rajho rastram). Here ‘king’ qualifies 
‘kingdom’. But, on this analogy, we cannot say: the hermit’s kingdom, 
even though the hermit and the kingdom are also different from each 
other. Similarly, we can say ‘the smell of sandal’, but we cannot say 
‘the smell of something having colour’, even though in both cases, there 
is difference between the two things. To come to cases where there 
is no difference, we can say: dmrdndm vanam = a grove of mango- 
trees. Here mango-trees qualify grove, though the grove is not 
different from the trees. That which qualifies must keep out something. 
When we say, the smell of sandal, the word sandal keeps out other 
things having smell like flowers. When we say, the king’s man, the 
word ‘king’s’ keeps out other possible masters. It is not, therefore, the 
fact of being different which enables qualifiers to qualify a substance. 
Sometimes, the whole qualifies a part as in ‘rcah padam ’ = the word 
of a stanza or grhasya dvarasald = the parlour of the house. Here 
the house qualifies the parlour, because parlour is part of the house. 
In other words, the relation of qualifier and qualified cannot prove 
that two things are different from each other.] 

35. The experts’ knowledge of the genuineness of 
precious stones and coins, uncommunicable to others, 
is born of practice and not of reasoning. 

Expert examiners of coins, goldsmiths etc., even after 
having found out the subtle reasons for deciding the genuine¬ 
ness of coins like a k5.rsa.pana, are not able to communicate 
them to others, because these reasons have no words to ex¬ 
press them. Even experts, full of concentration, do not 
understand, without long practice, the distinction between 

I. 36-37. 


sadja, rsabha , gandhara , dhaivata etc., even though, it comes 
within the range of perception. 1 

[1. The purpose of this verse seems to be to point ^out that, in 
order to understand certain things, it is necessary, not only to make 
use of perception, inference and tradition, but a fourth thing called, 
abliyasa, practice. The knowledge which the expert acquires about the 
genuineness of precious stones is something which he cannot communi¬ 
cate to others nor has it been communicated to him by others. That 
Is why it is said to be different from tradition which can be communi¬ 
cated to others.] 

36. The extraordinary powers of the Pitrs, the 
demons and the goblins, going beyond perception and 
inference, are the results of their previous deeds. 

In all circles of thought, it is admitted that deaf people 
and others hear sounds in their dreams and that some per¬ 
sons can see minute things inside houses through thick walls, 
without breaking them. When the question arises as to the 
means by which this is done, one cannot think of anything 
else than the indefinable power of adrsta. the result of deeds 
done in previous births. 1 

[1. Vr. points out that those who become deaf after birth may 
hear sounds in dreams because of the impressions of sounds heard 
before becoming deaf. But those who are born deaf also hear sounds 
in dreams. And then there are the manes, demons and goblins who 
can see the past, the distant and the hidden. All this is due to the 
effect of deeds done in previous birth. This faculty is called adrsta 
which must be distinguished from perception, inference, tradition and 

37. The knowledge of the past and the future of 
those whose insight has manifested itself and whose 
mind is in no way tainted differs in no way from 

According to the view that the effect is a totally new 
product ( asatkaryavada ) it is, before its creation, without 



a basis, without a form, without an essence, something 
in regard to which the causal power of something can¬ 
not be determined nor specified. How then can it be under¬ 
stood? According to the other view ( satkciryavdda ) its speci¬ 
fic form (before creation) is not manifest, it is unknown for 
the purpose of worldly transactions, it is as good as totally 
non-existent. But cultured persons, whose impurities have 
been burnt away by austerities, whose cognitions are free 
from all limitations, see everything vividly reflected in their 
cognitions. 1 

[1. Vr. points out that the purpose of this Karika is to state that 
in addition to ordinary perception, there exists another. Only those 
whose impurities have been washed away have it. It is the result of 
their good deeds in previous births. Through that, they can see the 
past and the present. 

While explaining the expression: adrstapratiniyatakdranasakti- 
•parigraham, Vr. points out that the different kinds of inference: purva- 
vat, sesavat and sdmtinyatorfrsta, cannot enable us to determine what 
has causal power in regard to which effect, no matter what view of 
causality we adopt, whether it be asatkdryavdda or satkdryavada .] 

38. The words of those who, with their divine 
vision, see things which are beyond the senses and un¬ 
knowable. cannot be set aside by reasoning. 

The supreme inner Controller, the atoms which are the 
abode of the initial cause of creation, the unmanifested Word 
—Absolute which is the substratum of its powers, the gods, 1 
the residual forces 2 generated by action, leading to particular 
results and not to others, in their maturity, the divine other¬ 
worldly body, 3 and other such things known in all scholarly 
circles, imperceptible to the senses like colour and beyond 
the range of inner experience like happiness are undoubtedly 
perceived by the sages with their extraordinary eye. The 
words of these sages, dealing with matters beyond the range 
of inference, cannot be upset by reasoning which is so liable 
to err. 4 People born blind cognise colour and as they had 
never seen colour before, it cannot be a case of inference. 

[1. Those who have a visible or concrete form and those who have 

2. Aniibandlia-parhmma-saktivaikcilyani. The residual force of 
action is anubandlia; when under favourable circumstances, the force 
yields the result of the action in question, it is parinama; its inability to 
yield any other result is sciktivaikalya. 

3. Siiksinamativaliikam sariram. Vr. says that this refers to the 
intermediary body which the dead person acquires before he gets his 
next body. It is called intermediary, because it enables the person to 
cross over. 

4. Vyabhicaribhir cinumanaih —This is a reference to the point 
already made in Ka. 32 and 34 that all reasoning is unreliable and 
liable to be upset by other reasoning.] 

39. How can one who does not question the 
authority of such persons any more than his own ex¬ 
perience and has, therefore, taken his stand on direct 
vision, be set aside by others (who follow reasoning) ? 

There are in every branch (of the Veda) and for every¬ 
body, some trustworthy persons. Their words are not scru¬ 
tinised and are not to be doubted any more than one’s own 
perception. Whose actions, like making stones float on water, 
though not actually seen, can be believed on the authority 
of trustworthy persons who follow them. It is like this. That 
actions done in this world lead to desirable or undesirable 
fruit after death is accepted in the world on the authority 
of trustworthy persons and even without such teaching in 
the Scripture, all persons usually accept it. 

40. In order to decide what is good and what is 
bad, all men, including the lowest, have very little use 
for scripture; 1 

Therefore, direct perception and extra-sensory cognition 
(of the sages) set aside reasoning even when there is con¬ 
tradiction between the two. 



[1. This verse is found in the SI. Var. (verse 3, p. 209. (Chow- 
khambha edition). It is there given as that of Parasarya, i.e., Vyasa. 
Jayanta Bhatta, in his Nyayamanjarl, also gives it as a quotation from 
Vyasa. Here, however, it is not given as a quotation, but as an integral 
part of the Vakyapadiya itself.] 

41. One who has recourse to Tradition which 
shines uninterruptedly like the T consciousness cannot 
be diverted therefrom by mere reasoning. 1 

The spontaneous and eternal consciousness which is 
attested by our sense of ‘I am’ and the like, is not set aside 
even by the teaching of authoritative persons that there is 
neither *1’ nor < mine’ even in the case of libeiated souls as 
far as worldly life is concerned. Such is the case with Tra¬ 
dition in the form of Scripture and written tradition accept¬ 
ed by all cultured people. It is not violated in the matter of 
what should be done and what should not be done, what can 
be eaten and what cannot be eaten, which woman one can 
marry and which woman one cannot many by the followers 
of different doctrines. One who closely follows such a tradi¬ 
tion observed by the elders cannot deviate from the right 
path by accepting the reasoning of logicians. Thus, he re¬ 
mains irreproachable in the eyes of the woild. 

[1. Mere reasoning such as that the word, the meaning and their 
mutual relation being eternal and known from the world, it is not 
necessary to compose the science of Grammar.] 

42. Fall is not unlikely in the case of one who 
relies on reasoning, as in the case of a blind man who 
walks along a difficult path by groping with the hands. 

Just as one, who, after seeing a sample, accepts the rest 
according to the maxim of the rice in the cooking pot (st/ia- 
llpulcikanydya ) 1 is like a blind man who goes hurriedly on 
a difficult hilly path without the help of one who can see, 

t 43. 51 

understands a part of the path by groping with his hands, 
traverses it, and, on that basis, takes the rest of the path also 
to be similar and comes to grief, in the same way, one who, 
without the help of the eye of Tradition, relies on reasoning 
and, after having attained confidence in some matters through 
inference only, proceeds, without the help of tradition, to 
perform acts having visible and invisible results, inevitably 
incurs great sin. 

[1. Sthdlipuldkanyaya. See Jacob — A Handful of Popular Maxims —1, 
p. 53.] 

43. Therefore, on the basis of eternal Scripture 
and authoritative tradition, the cultured have composed 
this science of the explanation of the words. 

Therefore, teachers have composed this science of the 
explanation of words (Grammar) after having taken as 
authority the tradition which is not connected with any per¬ 
son, not to be doubted and calculated to teach man what is 
good for him, after having accepted, as in the case of words 
like prsoclara 1 the uninterrupted practice of their own schools, 
acceptable to the cultured, in regard to the use of the cor¬ 
rect forms of words and after having taken as authority, the 
works of earlier sages in which option is allowed where there 
is contradiction between two teachings, in which there are 
main rules and their exceptions and in which there is great 
variation, according to time, in the manner of presentation 
of the forms of words . 2 

[1. Prsodara. See P. 6.3.109— Prsodarani yathopadistam. 

2. Pratikalam drstasabdasvarupavyabhicdrani Vrsabha under¬ 
stands this in two ways: (1) during periods when men are intellectually 
strong, everything is taught in great detail; at other times, briefly. 
This is one variation according to time. (2) At one time, one form 
of a word is taught as leading to happiness (merit) and another form 
at another time. At one time, nydnkavam was correct but now ?iai- 
yankavam is correct.] 


V 44 . In the words which are expressive the learn¬ 
ed discern two elements: one is the cause of the real 
word which, the other, is used to convey the meaning. 

That is called the expressive word which takes over the 
meaning, imposes it on its own form, makes it identical with 
itself, as it were. It has been said in the Sangraha. 

“According to the view that it is the undivided word 
which conveys the meaning , 1 the expressive word has its 
own form (which is the basis of its usage and grammatical 
explanation). According to the view that the word can be 
divided into meaningful parts , 2 the basis of the formation 
of the word is the element delimited by meaning. Others are 
of the view that the word which conveys the meaning (does 
not express it but only) suggests it, because in that way, 
the use of the relation of the identity of the word and the 
meaning in the form ‘It is this’ would be possible. 3 ” Or the 
word updddna may be taken to mean the whole which is to 
be accepted (for the operation of the rule). It is like this: 
where a word stands for its own form, its parts are not taken 
into consideration and no division is, therefore, cognised . 4 
What is meant by ‘one is the cause of the real word’ is: the 
cause of that, standing on which or resting on which or based 
on which the sounds convey the meaning. What is meant 
by ‘the other is used to convey the meaning’ is—By the acti¬ 
vity of the articulatory organs, it attains modification, be¬ 
comes audible, depends, as the illuminator, always on the 
illuminated 5 and is used to convey the meaning. According 
to others, that in which there is no inner sequence is the 
cause, that which has acquired sequence conveys the mean¬ 
ing. They proclaim that even that conveys the meaning 
when it enters the mind of the listener with all sequence 
suppressed. Another maintains that what has sequence is the 
cause of what has none. In the sequenceless word are merg¬ 
ed both the expressive power ( srutisakti ) and the power of 
being expressed ( arthasalcti ). It has been said— 


t 45. 

“From the differentiated, the undifferentiated word is 
born and it expresses the meaning. The word assumes the 
form of the meaning and enters into relation with it.” 

[1. Avyutpattipaksa. When a word cannot be divided into meaning¬ 
ful parts, the whole of it is associated with a meaning which is not 
related to the meanings of its parts, because there are no parts. 

2. Vyutpattipaksci. Where a word can be divided into meaningful 
parts, the meaning of the whole is sought to be connected with the 
meaning of the parts. 

3. Dyotaka. Those who hold that the word is suggestive (and not 
expressive = vacaka ) seem to mean that the relation of expressive and 
expressed presupposes difference between word and meaning. But 
there is abheda between the two. Therefore, the word cannot be vacaka , 
it can only be dyotaka . 

4. Sometimes, a word stands for its own form and not for a 
meaning different from the form. That is mostly the case in gram¬ 
matical rules. P. 4.2.33 “agner dhak” says that the suffix dhak (= eya) 
is to be added to agni. Here agni naturally stands for its own form 
and not for what it means. Therefore, it cannot be divided into parts. 
The suffix can be added only to the word as a whole and not to any 
part of it. Agni is svarupapaddrthaka . 

5. Pratydyyaparatantra . ‘Tadakdropagraheiiopasthandf —Vr. The 
word as prakasaka, naturally depends upon the meaning which is 
prakdsya or pratyayya . It is supposed always to carry the reflection 
of the meaning.] 

45. That there is an essential difference between 
them is the view of some followers of tradition. Others, 
on the other hand, think that they are one but appear 
to be different owing to difference in the point of view. 

When the word ‘essential difference’ is the object of the 
verb bru (ahull), it takes the second case-affix (dviiiya) but 
when the whole sentence as uttered by the followers of tra¬ 
dition is imitated, it takes the first case-affix (prathamd). 
According to the view that the effect is different from the 
cause, the difference between the two (kinds of words men¬ 
tioned in the previous verse) is clear. According to the 
other view (that the effect is not different from the cause) 


Vakyapadiyam of bhartrhari 

one and the same thing is thought of as two on the basis of 
difference in point of view resulting from its two powers. 
Or the difference of opinion among ancient thinkers may be 
based on their difference of opinion regarding the identity 
or difference of the universal and the particular of words . 1 
This matter will be discussed in detail later. 

[1. sabdakrtivyaktibhedabheda—“Tatra kesamcic chabdakrtir nimit - 
tam sabdavyaktir vacika . Aparesdm etad eva viparltam. Tayos ca kecid 
bhedam abhedam cdpare pratipannah. (Vr.)] 

46. Just as the fire which is within the churn- 
sticks is the cause of the other fire (which is kindled), 
similarly, the word which is in the mind (of the 
speaker) becomes the cause of the different expressive 

Just as one sees that, both according to the doctrine of 
identity and that of difference , 1 the fire, at first in the form 
of seed and unmanifested is later kindled into a flame and 
its capacity to illuminate itself and others, previously imper¬ 
ceptible, becomes later perceptible, in the same way, the 
word in the mind which has the seeds of differentiation in it, 
comes near to manifestation when the seeds become mature 
and is manifested by the movements of the articulatory 
organs and due to differentiation in the manifesting sounds, 
appears to have divisions and is perceived as having inner 
sequence and illuminates its own form and others . 2 ^ 

[1. Ekatvapaksa and nanatvapaksa stand for satkarycivdda and 
Qsatkaryavada respectively. 

2. The word exists in the mind of the speaker as a unity. When 
he utters it and produces the different soimds by the movements of 
the articulatory organs, it appears to have differentiation. But the 
listener ultimately perceives it as a unity and it is then that it conveys 
the meaning. I n this verse, the word which exists in the speaker’s 
mind as a unity is described as the nimitta . As Vr. puts it: Sa eva 
vaktrbuddhistho’nabhivyaJcto nimittam.] 


I. 47-48. 


47. First conceived in the mind and applied to 
some meaning or other, the word is suggested by the 
sounds whch are produced by the articulatory organs. 

When the relation of identity between the word and the 
meaning in the form ‘It is this’ ( so’yam) is analysed, the 
word which is sought to be superimposed on the meaning 
and into which the meaning enters as a kind of reflection 
(upayogenanupravesam labhate) is, before its utterance, visua¬ 
lised by the mind, applied to a meaning according to the inten¬ 
tion of the speaker, seems to change its form into that of 
another and projects it (on the meaning). What is meant 
by ‘produced by the articulatory organs’ is that the word 
whose nature it is not to change, appears to change, follow¬ 
ing the manifesting sound whose nature it is to change. 
The word which is entirely without change appears to change 
in terms of the gross nada, perceptible like a mass of clouds, 
brought into being when the subtle all-pervasive sounds 
. (dlivani ) are amassed (praciyarrvane) by the movements of 
the articulatory organs 1 . 

[1. Three words are used in the Vrtti-sabda, dhvani and nada. 
The first stands for the word which conveys the meaning. The 
dlwanis are conceived of as something atomic, all-pervasive and imper¬ 
ceptible. When amassed by the movements of the articulatory organs, 
they become gross and perceptible and are then called nada. It is 
they which suggest the word. The word is first conceived as a 
unity in the mind of the speaker. In order to utter it, he makes 
movements of his articulatory organs which have the effect of progres¬ 
sively collecting the subtle, atomic, all-pervading dhvanis and bringing 
into being the nadas which are gross and audible. These have divisions 
and sequence and so the word, suggested by them, though changeless 
and sequenceless, also seems to have them. The process of accumu¬ 
lation of the atomic dhvanis. resembling the accumulation of clouds 
(abhrasahghdtavad ) is referred to in Verse 111—** abhramva praciyante 
sabdakliyah paramanavah ”.] 

48, Because the gross sound (nada) is produced 
in a sequence, the word which is neither prior nor 
posterior nor has any sequence, is manifested as having 
sequence and parts. 



The gross sound ( nada ) which is in the form of an accu¬ 
mulation on account of the sequential activity (of the articu¬ 
latory organs) suggests the word ( spliota ) by means of the 
functions of prevention and permission . 1 Though the word 
(spliota) is one, it appears to have different parts. It has 
neither sequence nor simultaneity, both being opposed to its 
one-ness and eternality. Therefore without giving up its one¬ 
ness, it appears to have differentiation which is a property 
of the gross sound associated with it. Such is the nature of 
what is associated with it. It is like the whole ( avayavi ) 
which is an absolute unity appearing to be differentiated be¬ 
cause of the differentiation of its parts. 

[1. Sapratibandhdbhyanujhayd vrttyd. Pratibandha (prevention) 
and abliyanujha (permission) are usually given as the two functions 
of time, by virtue of which everything happens at its proper time, 
neither before nor later. Growth, for instance, means that at any of 
its stages, the later stages cannot appear. Similarly, when a word is 
uttered, the manifesting sounds can only appear in a sequence. When 
one appears, the later ones cannot. This is abhyanujna for what ap¬ 
pears and pratibandha for what does not.] 

49. Just as a reflection, found elsewhere (as in 
water) seems to have movement because of the move¬ 
ment of water, such is the relation between spliota and 

According to the view that the reflection is identical 
with what reflects it or according to the opposite view ( anya- 
tvapalcse) , the reflection of an object like the moon appears 
to be connected with and to have the properties of its sub¬ 
stratum, but it is not really so. Though it has no movements 
of its own, it takes on the property called movement of water, 
wave etc., and appears to have the various movements 
of water etc. Similarly, the word ( spliota ) takes on the pro¬ 
perties of the primary sound (prdkrtadhvani) when it is look¬ 
ed upon as short, long, or protracted and those of the secon¬ 
dary sound (vaikrtadhvani ) when it is looked upon as having 
quick, medium or slow speed . 1 

I. 50. 


[1. There are different views about the nature of a reflection: 
(1) In the presence of the original, a luminous object like water is 
affected and seems to reflect the original, but it is only a change in 
the water.. There is no reflection apart from the water. (2) The re¬ 
flection exists in the water apart from it and has the original as its 
material cause. (3) The rays of the eye, turned back by luminous 
surfaces like those of a mirror, water etc., see the original and that 
seeing is the reflection. According to the first view ( tattvapaksa ) the 
xeflection appears to be in the water and to have its movements, but 
it is only the water which one sees in that condition. In the second 
and third views (anyatvapaksa ), the reflection is something different 
ficm the water. There is no contact between the wind and the 
reflection and so there cannot be any movement in the reflection itself. 
Because the reflection is in contact with water and the latter with the 
wind, the movement in the water appears to exist in the reflection. 
According to all the three views, the reflection by itself is without 
movement, but appears to have it because of the water. Similarly, the 
word ( sphota ) appears to be short, long or protracted because' the 
sounds which manifest it ( prakrtadhvani ) are either short, long or 
protracted or it appears to be of quick, medium or slow speed, because 
the secondary sounds ( vaikrtadhvani ) are such.] 

50. Just as, in knowledge, its own form and that 
of the known are cognised, in the same way, in verbal 
cognition, the form of the object and that of the word 
are cognised. 

Knowledge is dependent on the object because it is there 
to illuminate it. Though it is not meant to show its form also, 
still it does show it separately as another knowledge would. 1 
That is how a previous knowledge, not cognised by another 
knowledge, becomes the object of remembrance. In the 
same way, the word, in so far as it is secondary to 
the meaning, is dependent upon it and assumes its form. 
It also conveys its own form, as a primary thing. Though 
it is present, it is not resorted to in the world, because 
it cannot have any relation with actions like eating. In Gram¬ 
mar, on the other hand, the external object is opposed to 
(grammatical) operations and the own form of the word 
appears as the meaning of a word and, therefore, there is ncr 
opposition between it and action, 

K. S. 8 



[1. Knowledge becomes known in two ways, according to the Vrtti. 
It can become the object of another knowledge as in the anuvyavasdya 
of the Naiyayikas which takes a form like: ghatam aham janami = I 
have the knowledge of the pot. But even in ordinary forms of know¬ 
ledge, when an outside object is illuminated, the knowledge itself is 
also revealed without its becoming the object of another knowledge. 
That is why we can remember a previous knowledge. Only that 
which had become the object of a previous knowledge can be 

51. The energy ( Kratu ) called the word, existing 
within, as the yolk in the pea-hen’s egg, has an action¬ 
like function and assumes the sequence of its parts. 

The external (audible) word, employed in verbal usage, 
is merged in the mind after suppressing all assumption of 
differentiation, without, however, abandoning the residual 
force of the differentiation, as in the case of the yo m the egg 
of the pea-hen. 1 Just as one single word can merge, so can 
passages consisting of as many as ten parts The word, thus 
merged, with all differentiation suppressed again assumes 
differentiation and sequence, when, through the speaker’s 
desire to say something, the inner word is awakened and it 
becomes the sentence or the word, each with its divisions. It 
is this merging and emerging of the word which is looked 
upon as its activity. 

ri Mayvrandarasavat —a comparison usually brought in to explain 
how what looks one and simple may actually contain all the potentia¬ 
lities of multiplicity and complexity. See K A. Subramama Iyer-Some 
more Nyayas (JOR., Madras, Vol. 6, p. 342). 

2 Dasataya. This word occurs in the Vrtti on verse 82 also. In 
both the passages, Vr. says that it means Catulisastih. See He. s com¬ 
mentary on Vak. C. Ka, 66, where the word occurs: Dasa avayava 
yesam te dasatayah samudayah rgdtmanah. Tesam samuhasya dasa- 
tayasya catuhscLstytitTiidkasya, etc.] 

52. Just as the unified image of an original figure 
is drawn on cloth (in three stages), so does one see the 
three stages in the case of the word also. 

I. 53. 


When a painter wishes to paint a figure having parts 
like that of a man, he first sees it gradually in a sequence, 
then as the object of a single cognition and then paints it 
on cloth or on a wall in a sequence. In the same way, the 
word in verbal usage is first perceived in a sequence, then 
cognised as a unity with the sequence suppressed. This part¬ 
less and sequenceless mental form is superimposed, i.e., iden¬ 
tified with the previous appearance having sequence and 
seeming to be separate. It again enters into verbal usage 
by displaying the characteristic of the sounds, namely, diff¬ 
erentiation and sequence, produced by the movements of 
the articulatory organs. In the same way, the word goes 
again and again through three stages and does not fail to 
become both the illuminator and the illuminated. 1 

[1.. It has been said in verse 44 that there are two kinds of words, 
of which one is the nimitta and the other, the expressive one. As to 
which is the nimitta and which the expressive one, is a question of 
point of view. Different points of view have been indicated in the 
■Vrtti on verse 44. Of the two kinds of word, one is a unity, with no 
differentiation and inner sequence and the other has differentiation 
and sequence. Either of the two can be looked upon as the nimitta, 
but only the word which is a unity and has no differentiation can be 

The three stages metioned in the verse and the Vrtti are described 
as follows by Vrsabha Prathamatah sravane kramavan , tato’l'pabhutah 
kramahy tatah parapratipadavdya kramavan itiJ] 

53. Just as the mind of the speaker first turns 
towards the words, in the same way, the attention of 
the hearers also is first directed towards them. 

Just as the speaker, wishing to make an effort to utter 
a particular word, concentrates his mind on it, isolating it, 
as it were, in each case, from all the other words around it, 1 in 
e same way, the hearer also, knowing that the understanding 
of the meaning depends upon a clear grasp of the form of 
the word, ascertains it as associated with all its attributes. 2 
This stage consisting in grasping the form of the word, is 



not thought of as a separate thing because of one’s habit of 
concentrating on the understanding of the meaning. 3 There¬ 
fore, all words capable of conveying all meaning, first attain 
supreme importance ( sesibhavakcLstliam ) and then become 
secondary to the conveying of the meaning. 

[1. Pratisabdam paritah paricchinndn sabddtmanah samsprsanniva — 
When one wants to utter a particular word, one has to grasp its form 
clearly, i.e., one has to separate it, as it were, from the other words 
which are also lying dormant in one’s mind. But the other words can¬ 
not all come up in his mind at that time. His mind does not touch 
them at that time ( samsprs ). That is why the word iva is used: 
samsprsann iva. At the most, the speaker would think of the other 
words in the sentence which he is about to utter, but not to the same 
extent as the word which he is actually pronouncing. In either case, 
the word iva is significant. 

2. Tam sabdam sarvaih sambandhibhir visesanair dsritasamsargd- 
nugraham. What the author means is: tarn sabdam asritasarvasam - 
bandhivisesanasamsargdnugraham . 

3. Arthapratipattibhavanabhydsat . According to Vr. this means: 
on account of one’s habit of understanding the meaning and acting 
accordingly. Bhavana kriyanusthdnam.'] 

54. Being meant for something else, the world 
does not pay attention to the forms of the words which, 
being primary in some cases only, are ordinarily 
secondary to the meaning. 

Just as, in an expression like ‘this is a white one (ayam 
suklah) where something having a quality is understood, 
connection of the quality with a particular action takes place 
only sometimes as taught elsewhere, in the same way, when 
the identification with the form of the word takes place as in 
the expression: ‘this thing which is the word gauh’ and when 
action is meant to be connected with it, its purpose is fulfilled 
and when, therefore, it becomes secondary to the object, 
people do not connect it with action as they do connect the 
object with it. 1 

[1. In order to show that the form of the word becomes connected 
VyitJj action only sometimes, the analogy of action is ^iven. statements 


i. 55. 

present a quality as connected with action only sometimes. There is 
a difference between suklam gayi pasya and suklam gam bhojaya. The 
former sentence presents the quality white also as connected with 
the action of seeing, suklam is a visesana of gam. But in the latter 
sentence, it is the object cow which is to be fed. It happens to be 
white, so the white cow is fed. But the quality 'white’ is not 
connected with the action of feeding. It is only upalaksana 
\ and not visesana. Similarly, where actions are performed on objects con¬ 
veyed by words, as usually happens in the world, the form of the 
word is not so connected. But a word conveys its form also and it must 
also have its scope for being connected with action. It is in the 
science of Grammar that the form of a word finds its full scope for 
being connected with action. Where the form is secondary to the 
object which it coveys, people do not connect it with action.] 

55. Just as light has two powers, that of being 
revealed and that of being the revealer, similarly, all 
words have two distinct powers. 

Pots and other such things are always the cognised and, 
therefore, at the time of their cognition, they do not bestow 
the slightest benefit, either on the sense or on the object, as 
an aid to their cognition. 1 In the same way, all the senses 
become the cause of the cognition of the object without them¬ 
selves being cognised. Light, on the other hand, being en¬ 
dowed with a form opposed to that of darkness, becomes the 
cause of cognition, as a helper, when its own form is cog¬ 
nised. In the same way, the word, grasped in its form 2 
which is distinct from that of other things and other words, 
with its special form which is fully cognised, illuminates the 
object to be cognised. These two powers of the words, that 
of being the revealer and that of being the revealed, are 
always identical (<a tmabliute ) but appear to be different. 

[1. Grdhyatvenaiva—The object is always the thing to be cognised 
(grahya ) in the process of its cognition and never the grdhaka. It 
is always the karma } the object, in the process of cognition. Though in 
Grammar, the karma is a Karaka, that does not make it a grdhaka , 
it is always grdliya. The object which is cognised does not bring about 
the cognition. It does not cause the utpatti of the cognition, even 
though it is the grahya. 



2. Svarupeimvadhriyarroanabhedah and parigrlutavisesasabdasva- 
rupah mean the same thing. Vr. points out that the former expression 
emphasises the distinction from other things and words while the latter 
stresses its character as the conveyor of the meaning.] 

That being so, 

56. No meaning is conveyed by words which have 
not themselves become the objects of knowledge. By 
their mere existence, even though not understood, they 
do not convey the meaning. 

If words could, without becoming primary as the thing to 
be conveyed, become secondary to the conveyance of the 
meaning, then, they would, by the mere fact of their pre¬ 
sence, whether cognised or not, without becoming the object 
of a cognition, convey their respective meanings. But they 
do not do so. Therefore, in the process of the words becom¬ 
ing secondary to the meaning, their assuming the form of 
the primary, is a necessary part. 

57. Therefore, when the own form of the word 
is not understood, one asks the speaker ‘what did you 
say?’ When an object is to be revealed, the form of the 
senses is not similarly perceived. 

People in the world, knowing that the understanding of 
the meaning is dependent upon the understanding of the 
form of the word, ask: what did you say? in order to under¬ 
stand the form of the words when, after the words have been 
uttered, their form is not grasped. The senses, on the other 
hand, become secondary to the understanding of the mean¬ 
ing, without being connected, even slightly, with the state 
of being primary, without their own form being cognised. 1 

[1. Se^bhdvarrmtrdsamsparsendsamsrstdni. Here matra seems to 
have the meaning of ‘slight’, little’. Ordinarily, in the Vrtti, it means 
‘a part’. The word occurs frequently in the Vrtti and has several 
shades of meaning. 

I. 58-59. 


Aparicclunnasvalaksaimni. Here svalaksana means the same thing 
as svai'upa. Ordinarily, in the philosophical literature, svalaksaiia, the 
particular, as used as opposed to samdnyalaksanci, the universal.] 

58. These two abstracted powers of words, treat¬ 
ed as different, become, without opposition, the cause 
of operations depending upon difference. 

Just as, both in the world and in the Science of Gram¬ 
mar, we perform, in regard to things within which difference 
has been artificially made by the mind on the basis of diffe¬ 
rence in point of view (nimitta) all the operations which 
depend on real difference, in the same way, in regard to words 
also, when one separates by abstraction ( apoddlvara ) their 
powers of being revealer and revealed all the operations such 
as the application of names like ‘designation’ (samjnd) and 
‘designated’ ( samjnin ) have been taught in the science of 

[1. In the world, when we say ‘a gold ring’, we are applying the 
words ‘gold’ and ‘ring’ to the same thing. When we think of the 
material with which it is made, we call it gold, when we think of its 
shape, we call it ring. This is what is called vyapadesivadbhciva. On 
account of this difference in point of view, we san say: suvarnasya 
ahguliyakam, ie. we can use the sixth case-affix ( sastlu ) which we 
ordinarily do only when there is real difference as in rdjnah purusah. 
The sutra svam ritpam sabdasydsabdasamjna (P. 1.1.68) tells us that, 
in the Science of Grammar, grammatical operations are ordinarily on 
the basis of this difference by artificial abstraction.] 

59. Just as words like Vrddhi, having conveyed 
their own forms enter into relation with their named, 
the phonemes conveyed by the contractions at, aic, 

In rules like 

“Yan comes in place of ik when a vowel follows” 1 where 
the names are different (from the named) it is not the verbal 
element ik which is the original nor the verbal element Yan 


vakyapadiyam of bhartrhari 

the substitute. It is between the named having a. different 
form conveyed by these verbal elements that the relations of 
original and substitute is taught in the Science of Grammar. 2 
Even where the named have the same form as the names, it 
is clear that such a relationship is understood. Just as words 
like Vrddhi , standing for their own form, meaningful with 
their meanings, wanting to convey other sounds with their 
form, enter into relation with d etc., varied by accent, nasa¬ 
lisation etc., conveyed by u t, aic etc., in the same way (the 
name conveys the named) even when the difference bet¬ 
ween the two is not easily perceived. 

[1. P. 6.1.77. 

2. Rupantarayuktiindm samjndndm. If ik and yan are names 
(samjrui) the sounds conveyed by them would be the named, samj- 
nin. Therefore, one would expect in the text sainjninam and not 

60. In the same way, the word agni (in the sutra 
P. 4.2.33) having first conveyed its own form agni, 
enters into the relation (of name and named) with the 
word agni (used in speech) and conveyed by the word 
agni (in the sutra). 

In the sutra 1 

“The own form of a word” etc. (svam rupam sabdasya ), 
the name and the named are separately mentioned. There, 
there are two words which are heard and which are 
meant to convey another, and the words which are to 
he conveyed are also two in number and they stand to 
each other in the relation (of name and named) and are 
(therefore) the object of grammatical teaching. Therefore, 
(the sutra in question), makes that word agni which has as its 
meaning the other word agni having the same form, the 
name of the other word agni conveyed by it having the same 
form. 2 

[1. P. 1.1.68. 

I. 61-62. 


2. The construction of the last sentence in the vrtti is rather 
awkward and contains repetitions. The sutra svam riipam sabdasya 
etc. teaches the relation of samjiid and samjnin between a word men¬ 
tioned in a sutra of Panini and the same word, having the same form, 
coming in actual usage. Though the samjiid, and the samjnin have the 
same form, they are mentioned as different from each other. The 
sutia in question refers to a word which may come in any sutra of 
Panini as the savijUd and the same word, as it may come in actual 
usage, as the samjnin. These two are called pratipddaka in the Vrtti. 
The word which actually comes in a sutra (as agni in agner dhak) and 
the same word when it comes in actual usage, are called pratiyamana. 
The sutra in question tells us that the word which actually comes in a 
sutia. is the name ( samjnd ) of the same word actually occurring in 

61. Whatever word is uttered does not neces¬ 
sarily become the object of grammatical operation. Its 
power to bring another similar word to the mind is not 
taken away. 

The reason for the statement which has been made will 
be given in the next verse. The word which conveys 
(pvatydyaka ) is uttered for the sake of another and it con¬ 
nects that other word for which it is uttered with grammati¬ 
cal operations, causes them to be performed on it. When the 
word which is conveyed ( pratydyya ) and exists in the 
mind (of the hearer) is uttered for the sake of illustration, 
its power to convey another similar word is not hampered. 
This is the nature of all words that are uttered 1 

[1. The point in this verse is that all words, when uttered, have 
the power to bring another similar word to the mind and the gramma¬ 
tical operation in question is done to the one that comes to the mind. 
That which is uttered is pratyayaka, something which conveys another 
and that which is conveyed is pratydyya. The latter can also be utter¬ 
ed in its turn and it would then convey another similar word.] 

62. When a word is uttered to convey a meaning, 
it is secondary to the latter and, so, is not connected 
with action. It is, with the meanings conveyed by the 
words that actions are connected, 

K.S. 9 



Just as words like ‘bring the cow’ or ‘eat the curds’ are 
secondary to the conveying of the meaning and do not become 
auxiliaries of the action in question, so are words secondary 
to other words, as there is no difference as far as dependence 
is concerned. From this one understands that all things which 
are conveyed by words, whether they are perceived by the 
eyes or the ears, become auxiliaries to the action. 

63. When, whatever is considered as the common 
property between the standard and the object of com¬ 
parison, itself figures in acts of comparison, some other 
common property, different from it, is adopted. 

The standard of comparison, the object of comparison 
and the common property, this triad is well-established. In 
the sentence “the ksattriya studies like the brahmana”, the 
common property is mentioned in regard to the standard of 
comparison also. When the statement is ‘the study of the 
ksattriya is similar to that of the brahmana, the two students 
are mentioned as connected with the standard and objects 
of comparison and then, excellence etc., are understood as 
common properties of the two ways of studying, different 
from each other, because they exist in different substrata. 
Completion ( parinis'patti ) etc. would be understood as com¬ 
mon properties of excellence etc. connected with study. In 
this way, there is no limit to the process of difference (by 
abstraction) A 

[1. There are different grades in the process of comparison, we 
can say: (1) Bmhmanavad adlute ksattriyali (2) Brahinwrmdhyaya - 
nena tulyam ksattriyadhyayanam (3) Brahmanddhyayanasausthavena 
tulyam ksattriyadhyayanasausthavam. 

In No. 1, the common point is that both the brahmana and the 
ksattriya study; in No. 2, the common point is that their studies 
have" excellence; in No. 3, the common point is that their studies lead 
to completion. What was a common point in the previous comparison 
becomes the standard and the object of comparison in the succee ‘ng 
one. Study, excellence and completion are really one, but are looked 
upon as different by abstraction.] 

1 . 64 . 6 ? 

64. When an attribute which is the cause of 
degree is presented as an independent thing, degree in 
it can be known only through another attribute present 
in it. 

Whatever is presented as the main thing in the form ‘this’ 
or ‘that’ is substance. There cannot be a greater or lesser 
degree of it; therefore, when degree is sought to be expressed 
in an object, it is done through attributes (nimitta) which 
exist in it, differentiate it, are dependent upon it, are physi¬ 
cally attached to it and are active in bringing about degree. 
In the sentence: This is an excellent white thing ( prakrstali 
suklali ) even though the colour white distinguishes the thing, 
it is also active in bringing about degree and, therefore, that 
thing which has that quality and would otherwise have no 
degree, now attains degree. In the sentence: ‘the colour of 
this is white’ where colour is presented as a substance, degree 
is brought about by an attribute ( nimitta ) present in the 
colour. The universal ‘whiteness’ which inheres in white 
is one and cannot be a cause of differentiation and so diff¬ 
erentiation in the one having shades of difference within 
(avantarasya ) is brought about by an attribute present in 
the colour. Or it can be put in this way: As there is no 
word expressive of such attributes (in suklataram vupawn 
asya ) and as they cannot be conveyed by the abstract suffix 
(bhdvapratyaya ) they are understood from the word, like 
whiteness itself, and are understood as the cause of degree. 
As long as it is sought to bring about degree in anything 
which is presented as the main thing, in the form ‘this’ or 
‘that’, so long would there be no end to this process of 
abstracting some attribute or other. 1 

[1. Verse 61 makes the statement that a grammatical operation is 
not done on a word which is actually mentioned in a sutra but on a 
similar word occurring in worldly usage. The two words are different 
but are looked upon as the same because of resemblance. Verse 62 
gives the reason for the statement made in 61: what is mentioned 
is subordinate to something else and so is not subject to grammatical 
operations. Verses 63 and 64 give illustrations. The former tells us 
that what was the common point in the previous comparison becomes 



the standard and the object of comparison in the succeeding one by 
a process of abstraction while the latter tells us that by the same 
process, what is presented as a guna in one statement can be presented 
as a dravya in another. 

When a quality which is the cause of degree in a substance is 
presented independently, i.e., as a substance, it is only by another 
quality present in it that degree can be understood. In f suklatarah 
patah’j pa ta is the substance and degree in it is expressed by suklatara 
which denotes the quality of being whiter. But in suklataram rupam 
asya , colour itself is presented as a substance and not as a quality of 
cloth. But it is presented as having degree and this degree comes 
through some other quality like bhdsvaratva, brightness present in the 
white. For grammarians, whatever a word presents as the main thing 
to be qualified is substance ( dravya ). They do not follow the Vaisesika 
definition of dravya (Vai. Su. 1.1.15.). Even what is called quality 
can be so presented by words in which case it will have to be qualified 
by some other quality. In order that a quality may bring about 
degree in a substance it must be: (1) bhedahetu, (2) asrita , (3) sarrisrsta, 
(4) prakarse savydpdra. Due to the absence of one or more of these 
conditions in the sentence purusasya svdmi, purusa cannot bring abqut 
degree in svami . For the same reason, jati or the universal, cannot 
bring about degree.] 

That being so, in an utterance. 

65. When the word considered to be the meaning 
is itself uttered, it assumes a form quite different from 
the one (which it had as a meaning). 

After having given a reason (verse 62) and examples 
(verses 63, 64). For establishing difference, the considera¬ 
tion of the topic under discussion, i.e., the form of the word, 
is now being concluded. Whenever the word which is the 
meaning is uttered for the sake of illustration, on every 
such occasion, another form of it, the cause or the basis of 
it, would be distinguished, for such is the nature of anything 
that is uttered. According to some, it is the same word which 
is repeated as the conveyor—form ( abhidlvdna ) without its 
losing its character as the form-conveyed (cibhidheyatva ). 
It has been said in the Sangraha — 

“The form of the word (as the conveyed) never comes 
Under the organs of articulation ( karana ) anymore than 


I. 66. 

objects like a cow (which are also conveyed by words); it is 
always the conveyed; when the conveyor-form comes under 
the organs of articulation, the conveyed which has the same 
form appears to be uttered, even though it does not come 
under the organs of articulation. 1 

[1. The point which is emphasised from verse 61 onwards is that 
the same word can be looked upon both as pratydyaka and pratyayya. 
The difference would then be due to an abstraction made by the mind. 
Ordinarily, it is outside objects which are conveyed ( pratyayya ) by 
a word. Sometimes, as in Grammar, the word itself is the pratyayya. 
The word which is pratydyaka is uttered, but not the one which is 
pratyayya. As soon as the latter is uttered for the sake of illustration, 
it becomes pratydyaka in relation to another similar word which would 
then become the pratyayya .] 

66. Before the name becomes connected with the 
named, the name is connected with its own form as the 
meaning (the named). That is how it becomes the 
occasion for the use of the genitive and the nominative 

Even though a word may not have any other thing to 
be conveyed, it is not without its own form (as the con¬ 
veyed). On the basis of its own form ( svarupadhistlvdnam ) 
and after making it secondary, the word conveys some other 
meaning. As long as the name is not connected with the 
named, the latter is not its meaning and if it has no other 
meaning, it would not get the name of stem (pi-utipadika) 
and so no case-affix could be added to it. As words are ex¬ 
pressive (vdcaka) because of their difference (in regard to 
the expressed), in the present case, the meaning of the stem 
would not be different from the stem itself. 1 

[1. In the sutra Vrddhir ad aic (P. 1.1.1.) the word VYddJii is the 
name and the vowels at and aic are the named. Even before a date 
was written, the word Vrddhi had already got the first case-affix. It 
must, therefore, have been a stem (.pratipadika ). But nothing can be 
a stem if it has no meaning and Vrddhi gets its meaning only after it 
is connected with ddoic. How then did it become pratipadika? The 



answer is that its own form was already there as the meaning. So 
it became pratipadika and the first case-affix was added to it.] 

67. Because it has its own form as the meaning, 
the nominative case comes after the name. When it 
has an outside object as the meaning, difference comes 
in and is conveyed by the genitive case in the form 

When the form of the word is sought to be imposed on 
external objects, then they are meaningful with their own 
form which is their basis ( adhistlidna ) and the first case-affix 
is taught after them. When one says: “This is so and so”, 
the relation (of the name) with the named is specified, its 
expressive power is determined. As when one says: “This 
VS-luka is a cow” or “This lad is a lion”. The cause of the 
addition of the sixth case-affix to the named, namely, the 
absence of the meaning of the stem, is brought about by the 
connection of the name which stands for its form with the 
words which are the named. It has been said— 

“Since, whether in the form ‘that is so and so’ or in the 
form ‘this is the name of so and so’, it is with a meaningful 
name and not with a meaningless one that connection takes 
place; therefore, the relation between word and meaning 
is eternal.” 

68-69. According to some, in the sutra—“ svam- 
rupam” etc., the particular form of the word ( vydkti) 
is taught as the name ( samjnd ) of the form in general 
(jati). The general form as existing in a particular 
one undergoes grammatical operations. 

Others, on the other hand, look, upon the parti¬ 
cular form to be understood as the named in this sutra; 
in particular places, the particular form conveyed by 

I. 68-69. 


the general form, comes up (for grammatical opera¬ 
tions) . 

(While explaining the sutra svam rupam sabdasya etc., 
P. 1.1.68.) some commentators say:—“the own form of a 
word is the expressor, the illuminator, the conveyor of the 
word,” while others say—‘the own form of a word is the 
expressed, the illuminated, the conveyed.’ These two views 
have been set forth by these two verses. In regard to the 
universal, there are two views current among thinkers. 1 Some 
maintain that the individuals have their special forms. The 
form of the individuals is not something unknowable, un- 
nameable and non-existent. It is the individual cow which 
is the cow and not cowness. It is the individual blue which 
is the blue and not the universal blue. The universal is the 
cause of the uniformity of cognition. It is the same thing 
in regard to the different individuals in which there is no 
inherence (of the universal). It is this (uniform cognition) 
which is the cause of the inference of the existence of the 

Others, however, are of the opinion: words attain their 
form in regard to the universal as identical with the othei- 
wise inexpressible individual. Everywhere, from a thing in 
which the cause ( nimitta ) exists, a cognition corresponding 
to that cause arises. In such cases, those causes which have 
their own expressive words and those which have not, 2 bring 
about the same cognition and the application of the same 
word, on the basis of a partial or complete resemblance. That 
being so, in the sutra in question, the words ‘own form 
(svam rupam) stand, according to some, for the universal 
(jati) and in the expression ‘of the word’ ( sabdasya ) ‘word 
stands for the individual. According to others, it is just the 
opposite. Either the individual is the name of the universal 
or the universal is the name of the individual. This universal, 
(of the word) uttered in the different contexts, conveys the 
individual as connected with the relevant action (i.e. gram¬ 
matical operation). The universal is never uttered as un¬ 
connected with the individual. Nor is the individual ever 



brought into use unconnected with the universal. The speak¬ 
ers’ intention differs according to his purpose. One of the 
two is the main thing and the other comes inevitably with 
it. That it is the universal on which- the action is to be 
done according to the Science of Grammar or that it is the 
individual that is so is only a difference in the statement (in 
particular cases). 

In regard to the sutra svam rupam sabdasya (P. 1.1.68.) 
the tradition of the circles of Grammarians is very varied. 
It may be stated as follows—The mere form which is a part 
of that meaningful combination of form and meaning, posses¬ 
sed of the powers of the universal, particular etc. and referred 
to as the word ( sabdasya ) is the name. Or the meaningful 
combination including the form, characterised by the powers 
inherent in it, is the name. The fact is that what is called 
the word is a combination of parts. Form etc. are its parts. 
One does speak of a whole and its parts as different from 
each other as when we say ‘the branch of a tree’. 

Another declares: Even though an object remains the 
same, the meaning of the word (which refers to it) varies. 
A word makes a choice between the different powers relating 
to the same object. 3 For example, in the expression ‘this 
stick the pronoun presents the object as something existing 
before one’s eyes, not as something having the universal 
stickness’ even though that is also perceived. The pronoun 
cannot express the connection with the universal stickness. 
The power of the word ‘stick’ which expresses a particular 
universal is debarred, even though it is visible, from present¬ 
ing it as present before the eyes. In the same way, in the 
expression ‘the own form’ (svam rupam ), the words ‘agni’ 
etc., even though they have the universal in them, are present¬ 
ed as the words agni etc, not as universals of the words. By 
word in ‘of the word’, what is conveyed is the connection with 
the universal (of the word agni) and not its aspect as being 
the word agni. All this is just an illustration of the principle 
that a word, in one of its aspects, considered as a name, can 
convey the same word in another of its aspects as the named: 4 
All the different views relating to the sutra svam rupam etc, 

I. 68-69. 


have not been set forth in order to avoid prolixity resulting 
from one thing leading to another. 

[1. So far, the sutra svam rupam was explained on the basis of 
the view that the own form of a word that may occur in a sutra is 
the name and the external object which it denotes is the named. The 
present two verses set forth two other views in regard to the name 
and the named. They are—(1) that the individual aspect of a word 
is the name (samjhd) and its universal aspect, the named ( samjnin), 
(2) that it is vice-versa. As this involves the concept of the universals, 
the author refers to two views in regard to universals current among 
thinkers of his day. The first view is that the existence of the 
universal is inferred from our uniformity of cognition in regard to 
the individuals. The individuals themselves can be perceived and 
spoken about without the universal existing in them being directly 
perceived. There are other distinguishing features in the individuals 
which enable us to perceive them without the help of the universal. 
That there is uniformity in our cognition of the individuals does not 
mean that we cognise directly the universal existing in them. We 
only infer its existence on the basis of the uniformity of cognition. 
The second view is that words primarily convey the universal. They 
convey the individual also as identical with the universal and, therefore, 
nameable. The universal existing in the individual is what enables 
us to cognise the individual as such: it is the ‘nimitta’ and our cogni¬ 
tion of the individual takes the form of this nimitta according to the 
principle: sarvatraiva hi nimittannimittavatyartlie nimitta—svamipah 
pratyaya utpadyate. It is this nimitta which causes uniformity of 
cognition in regard to the individuals and the application of the same 
word to them. 

2. Drstabhidhanani adrstabliidhandni ca nimittani. Nimitta is of 
two kinds: kdraka and jnapaka. It is the latter which is subdivided 
into drstabhidlvana and adrstabhidhana. Gdtva is a jhapakanimitta 
which exists in the cow. It is drstabhidlvana , because it has a recog¬ 
nised word to express it. When we see different individuals having 
gotva in them, we see the same gotva (jati ) in all of them, our cognition 
(pratyaya ) of all of them is uniform and we apply the same word to 
all of them. As Vr. puts it, three things persist. Drstabhidhanesu tra- 
yam anuvartate jdtih sabdali pratyaya iti. utpalagandlia is given as an 
example of adrstabhidhana , because the word does not denote any 
universal, but only two things connected by a relation. In this case, 
therefore, only two things persist: the word and the cognition. As 
though not satisfied with this example, Vr. gives that of rajapurusa. 

3. Sabdo liyekavastuvisayandmapi saktirudm avaccliedenopagrahe 
vartate. Any given thing has many aspects. Words have the power 

K. S. 10 


, „„ _ given occasion, conveys one 
to convey them. But a given woid, on g __ this stick, ay am 

of them to the exclusion of others. In aya^d ^ eyes tQ the 

conveys the aspect of the stick being pi stick, 

exclusion of the universal which is also presen 

4. Tatra sabdantararthah sabdantararthasya ^ 

samjnatvenopddtyanvanasya samjnibhavam pratipacya . c 

is that sabdavyakti is samjna and sabdajdti is samjnm To«fa to 
both of them as sabdantararthah is using rather peculiar- language, a 
general characteristic of the Vrtti. It seems to be a bahuvrihi compound 
the analysis of which would be: sabddntaram (vyakurupam) jatnuparr 
va artho yasya sa sabdah sabdtintararthal}.] 

70. Some consider the word to be one whether it 
is accepted as transitory or eternal; others consider it 
as many, whether it is accepted as transitory oi eternal. 

According to those who believe that the word is one, 
there cannot he this distinction between the universal and 
the individuals (because that presupposes difference between 
the individuals); that is why this new point (the unity or 
the plurality of the word) has been taken up after expounding 
the point about the universal. According to the view that 
the word is eternal, its unity is the main doctrine. According 
to the view that the word is an effect (and so not eternal), 
the invariable cognition of sameness which arises when a 
phoneme or a word, once uttered, is uttered again, leads to 
the postulation that it is one. It is on the basis of the view 
that it is one that it has been said m the Vartti a 

“That is already settled, because the phoneme ‘a ’’is one”. 1 
It is between the cognitions of the same phoneme t at there 
is an interval of time or of other phonemes and not that the 
phoneme itself is different. That one hears the same word 
at different places is like the cognition of Being ( sattd ) oi 
the universal at different places or the form (of moon etc.) 
in water etc. Even according to the view that the word is 
different and that it is an effect, one must necessarily accept 
a kind of secondary one-ness when a word is uttered again 
and again. Whether the word be eternal or an effect, those 


i. 1L 

who are for plurality believe that a word having many mean¬ 
ings and phonemes found in different words is not at all the 
same unit. 2 

[1. Va 5 on the sutra ‘a-i-im’ (M. Bha. I, p. 16). 

2. In regard 1o the word, there are the three following pairs of al¬ 
ternative views held by different thinkers: (1) that it stands for the 
universal or for the individual, (2) that it remains the same even when 
it has another meaning or that it is a different word for every new 
meaning (3) that it is eternal or that it is an effect. These pairs of 
views are related to one another. The concept of the word as a uni¬ 
versal presupposes the existence of individual words in each of which 
the universal inheres. So, according to the view that the word is 
one under all circumstances, the concept of universal and particular 
cannot arise. There is a very close connection between the view that 
the word is eternal and that it is one. This is the primary concept 
of unity. But some kind of unity has to be postulated even according 
to the view that there are as many words as there are circumstances 
of its use because, in all these circumstances, we do recognise the word 
to be the same. This is the secondary kind of unity. Grammarians 
accept eternality, unity and both the concepts of the universal and 
the individual. 

71. Even when the words where they occur are 
different, the identity of the phonemes is not affected. 
Similarly, a word occurring in different sentences is 
one and the same. 

In the different words asva, arka, artha, it is the same 
phoneme a that is used in the world; it is perceived after 
an interval of time or after the intervention of other phonemes 1 
or not perceived at all when the causes of manifestation 
( nimitta ) are absent; it appears to be different when the 
different speakers are at different places and, hence, the mani¬ 
festing agents ( nimitta ) are different, just as the reflection 
of the same thing in a shadow or mirror or water. Similarly, 
a word abstracted from sentences and having different mean¬ 
ings like the word go or aksi is really the same word as long 
as the form is the same. 2 Even though as expressive of two 
different meanings, a word may be a noun or a verb, still 
it is looked upon as the same word. 



[1. Va 11 and 12 on the sutra ‘a-i-un 3 (M. Bha., p. 18.) 

2. Vr. points out that the word aksi, when taken as a verb can be 
derived from the root aks or as or ad. He derives asva as a verb fiom 
the root sin. Of course, ordinarily, these words are nouns.] 

72. There is no word over and above the phonemes 
nor is there any sentence over and above the phonemes 
and words. 

The phonemes which are produced in a sequence, which 
pei'ish as soon as they are uttered, which do not exist at the 
sametime and have themselves parts, cannot produce the word 
as distinct (from themselves); therefore, the word is nothing 
more than the phonemes. As the phonemes have parts and 
as that process of division can be continued to the point of 
impracticability (a vya^vahdrcLvicclieddt) there lesluts some¬ 
thing which is the fourth of the fourth, unnameable and 
beyond practicability (vyavahdv&tita) but no phoneme oi 
word. And if there is no phoneme or word, there cannot 
be a sentence as distinct from them. That is why it has 
been saidb 

“All this would result in non-eternality,”. 1 

As there cannot be a collection (of phonemes or of 
words) there would be no unit associated with a definite mean¬ 
ing (uyagrhl>t&rthah) nor any form fully delimited, (upacp- 
JUtah) . 2 

[1. Cf. M. Bha. on Va. 12 on the sutra a-i-uiy (M. Bha. I., p. 18.) 

2. It was said in the previous verse that a phoneme occurring 
in different words is the same phoneme and that a word occurring 
in different sentences is the same word. This means that there is 
no sentence over and above the phonemes and words. The phonemes 
cannot produce the word because they do not co-exist. Secondly? 
they themselves can be thought of as having parts and if one goes 
on dividing these parts, one comes to something which is infinitesimal 
and unnameable, the fourth of a fourth ( turiyaturlya ) as it has been 
called to show that it is something very minute. Such minute parts 
cannot produce the phoneme and so the phonemes cannot produce the 
word. If the phonemes and the words cannot be produced, the sen- 

I. ?3-74. 


tence, as something over and above them, does not exist, nor can it 
be produced by them. It would be something eternal and indivisible. 
It would also be impossible to conceive of words having fully delimited 
forms or well-defined meanings.] 

Another has said— 

73. There are no phonemes in the word nor are 
there parts in the phoneme. There is no absolute 
difference of the words from the sentence. 

As the effort relating to the whole as a collection of 
phonemes is different and, therefore, the sounds uttered to 
suggest the word are also different, there takes place, from 
the different cognitions having the appearance of the cogni¬ 
tions of the real phonemes, a perception of the word caused 
by means of the perception of parts in the form of phonemes. 
But it is really without sequence, without before and after, 
one, eternal, indivisible and it appears as a unit produced 
by smaller elements, the phonemes. But the further divi¬ 
sions of the phonemes like a, indescribable, beyond usage, 
considered different, are really fictitious. Therefore, their 
clearly intelligible unity is well-known in the usage of the 
Science of Grammar. As it is the sentence, whether con¬ 
sisting of one word or of many words, which is used to ex¬ 
press the meaning, there arises a perception of the word as 
means to the perception of the sentence. In regard to the 
sentence, there arise cognitions having sequence and having 
the form of phonemes and words, but they are not identified 
with the sentence. Therefore, phonemes and words have no 
separate existence from the sentence so defined, unbreakable, 

74. The practice (of the Grammarians) has been 
based on these different views. What the main view is 
to some is, to others, the reverse. 

In regard to the supreme, sequenceless, inner principle 
of the word, scholars, on the basis of different traditions re- 


VakyApadiyAm bF bharurhAri 

garding every topic ( pratyadhikaranam ) have resorted to diff¬ 
erent doctrines in following the usage of the Science of 
Grammar. It is as follows—If the audible form is the same, 
it is the same word, even when the meaning varies; if the 
meaning varies, even when the audible form is the same, 
it is a different word. For some, the difference is secondary 
and one-ness is primary. For others, difference is primary 
and one-ness is a matter of usage. It has been said— 

“The same word has many meanings—For example, 
aksdh 9 padah, m)dsdsh. >n 

Similarly, after having said— 

“The word ‘village’ (grama) has many meanings .” 2 
he (the Bhasyakara) concludes specifically by saying— 

“That word ‘village’ is taken which stands for the whole, 
including forests, boundaries and pasture-land .” 3 

[1. M. Bha on Va. 9 on P. 1.2.45. (M. Bha. I, p. 220). 

.. Aksa means, among other things, (1) an axle, axis (2) a gambling 
die (3) an organ of sense; jpada means (1) foot, (2) a quarter, (3) a 
weight; masa means (1) black gram, (2) a weight, (3) a spot on the 

2. M. Bha. I, p. 59, 1. 20. 

3. M. Bha. I, p. 59, 1. 23.] 

75. Of the sphota which has no temporal distinc¬ 
tion and which appears to follow the duration of the 
sounds, distinctions in duration ( vrtti ) have been 
declared on the basis of distinctions in the manifesting 

As the principle word-self is eternal, the function of the 
measure called ‘time’ does not affect in the slightest degree 
the duration of sphota. Inasmuch as the form of the sphota 
is perceived as mixed up with the sounds, therefore, the 
duration of the sounds appears to be that of the perception 
of sphota. Through the limiting factor of such a perception 
of sphota, involving difference in time, divisions in the form 
of quick, medium and slow speeds, each faster than the fol- 

I. 76. 


lowing one by one third , 1 are associated with the sphota 
(though they belong to the sounds). 

[1. Triblvdgotkarsena. It is not clear whether this means ‘greater 
by one third’ or ‘multiplied by three’.) 

If that is so, in the case of the short, long and protracted 
vowels also, it is the sound which would be the cause of the 
difference in time. Therefore, just as the long and the pro¬ 
tracted vowels would be called by the same name, the sutra— 
“A vowel followed by the symbolic letter t stands for one of 
that duration only ” 1 would be applicable to them. Similarly, 
in the case of the different speeds also, it would apply and 
there would also be the objection contained in the following— 

“If the symbolic letter t is added after the quick one, 
it should be added after the medium and the slow ones also, 
because of difference in time .” 2 To this, the following answer 
is given— 

76. According to the view that the word is 
eternal, the short, the long and the protracted being 
different in nature, distinctions in time belonging to the 
primary sound are attributed to the sphota. 

The sound is of two kinds: primary and secondary. The 
primary one is that without which the form of the sphota, 
being unmanifested, would remain unperceived. And the 
secondary one is that by means of which the form of the 
sphota is perceived again and again without interruption. 
The author of the Sahgraha says, indeed, as follows— 

“The primary sound is the cause of the perception of the 
word ( sphota ); the secondary one becomes the cause of the 
difference in the continuity of perception ( sthiiibheda ),” 

In the same way, out of the sentences uttered by Garga and 
others, of equal length because of equal number of parts some, 
on account of their special arrangement ( sannivesavisesa ) are 
perceived after a large number of repetitions while others 
are grasped after a lesser number of repetitions. Similarly, 



the short vowel, by its special nature, is perceived by a sound 
of lesser duration. With so much of manifesting cause, a 
knowledge which grasps its form is engendered. The long 
vowel, on the other hand, is manifested by a sound of longer 
duration. The protracted vowel is perceived by a sound of 
still longer duration. Because of the non-perception of diff¬ 
erence, the duration of the primary sound is attributed to the 
sphota and is looked upon, by courtesy, as the duration of the 
sphota in the Science of Grammar. 

[1. P. 1.1.70. 

2. Va 4 on P. 1.1.70.] 

77. After the manifestation of the sphota, the 
secondary sounds cause difference in speed of utterance 
but the essence of the sphota is not affected by them. 

Just as light, as soon as it appears, becomes the cause 
of the perception of a jar etc. and if it continues, becomes the 
cause of the continued perception of it, in the same way, 
the sound which continues after the manifestation of the 
word, causes the continuance of the cognition of the word 
by adding strength to the manifestation. Therefore, though 
associated with the secondary sound, the difference of which 
is clearly perceived , 1 no identity is superimposed on the sphota 
which, therefore, does not cause any usage of difference of 
time in the Science of Grammar as in the case of short etc. 

[1. upalaksitavyatirekena vaikrtena dhvanina. Vr points out that 
the primary sound causes the manifestation of the sphota and so it 
has to be produced before the manifestation whereas the secondary 
sound is produced after the manifestation and so its difference from 
the sphota is clearly perceived by the hearer. There does not, therefore, 
take place the false attribution of its properties to the sphota. The 
two are clearly distinguished. As the manifestation of the sphota itself 
depends upon the primary sound which has to be produced before 
its properties are falsely attributed to it.] 

How does the sound become a factor in the cognition of 
the word? 

1. 78. 


78. Those who hold that the word is manifested 
by the sounds conceive of it in three ways—that a 
certain fitness is brought about in the sense of hearing, 
that a fitness is produced in the word itself or that a 
fitness results in both. 

In this verse, only the process of the manifestation of the 
word is stated. Giving illustrations is the topic of the next 
two verses. 

In this matter, some think that the sound, when pro¬ 
duced, perfects the sense of hearing and the sense of hearing, 
thus perfected, becomes the means of the perception of the 
word. 1 Other upholders of manifestation are of the view that 
the word, perfected by contact with the sound, becomes the 
object of the sense of hearing. According to others, the 
sound brings benefit both to the word and to the sense of 
hearing. The sense and the object, favoured by a helping 
factor, cause the perception of the word. Like the benefit 
conferred by the light of the lamp in regard to things per¬ 
ceived by the sense of vision, (here also) the powers of the 
causal factors, merely because they require a helping factor, 
do not go beyond the totality of the causal factors. 2 

[1. Vrsabha understands the perfection of the sense of hearing itself 
in three ways: (1) the sounds impart to the sense of hearing the 
power of grasping the word which was not there before. (2) the 
sounds only awaken the power which was already there (3) what 
is called perfection of the sense means the presence of the sounds in 
the sense of hearing, favourable to its grasping the word. 

2. What is meant here is that the function of the sounds, whether 
it be the perfection of the sense of hearing or of the word, does not 
go beyond the totality of the causal factors, but is part of them, just 
as the light of the lamp is part of the causal factors in the perception 
of visual objects. 

The Vrtti here may be compared with the following passage: — 

Sa ca nadah srotrasydnugrahe varttate. Tadanugrliitam srotram sab - 
dopalabdhau samartham bliavati, yatha aksno ranjanam ityeke. Apare 
sabdasyaiva, yatha proksanam prthivyd eva , na ghrdnasya. Ubhayor 
ityapare, yatha vasyendriyasya prathamatascaksurasviln ghatam cdnu - 
gyhnati . 

K.S. 11 



(Bhartrhari’s commentary on the Mahabhasya, p. 20 (A.B.O.R.T, 

79. By concentration and by the application of 
ointment, it is only the visual sense which is perfected. 
The perfection of the object is for the perception of the 
smell in it. 

Concentration, whether natural or extraordinary, does 
not bring about any special feature in the essense of the ob¬ 
ject to be seen. Similarly, substances like ointment perfect 
the sense of vision and not the external object. Even extra¬ 
ordinary concentration confers benefit on the sense of vision 
only in the perception of things subtle, hidden and distant. 
If it conferred a benefit on the object, others also would per¬ 
ceive it, as there would be no difference. Perfection of the 
object, ie. of oil etc. 1 by sunshine and of the earth etc. by the 
sprinkling of water, can be seen at the time of the cognition 
of their smell, but not perception of the sense of smell. If it is 
the sense of smell which is perfected, there would be no 
difference as far as the cognition of the smell is concerned, 
whether the objects be perfected or not, because there would 
be no difference (in the objects). 

[1. Tailddindm, Vr explains: Gandhatailadindm = of medicated 
oil etc.] 

80. According to the view that the visual sense 
reaches out to the object, it is held that both the object 
and the sense are perfected by light. Such is the process 
in the case of the (manifesting) sounds. 

One who is situated in the dark perceives an object like 
the jar, illuminated by light. According to those who believe 
that the sense of vision does not reach out to the object, it is 
the object which is mostly perfected. If the sense of vision 
reaches out, the rays of the eye are aided by light which is 
of the same kind. 1 

I. 81 


[1. There were several views about the sense of vision. Some held 
that it is nothing more than the physical eye-ball and, therefore, not 
of the nature of light. Others held that it consists of the eye-ball 
which, however, is of the nature of light. The Naiyayikas held that 
it consists of visual light rays and, therefore, it is taijasa. Even here, 
there were two views: Some held that it does not reach out to the 
object before perceiving it, that it is aprapyakdri. Others, however, 
held that it reaches out to the object ( pidpyakdri ) Vr describes the 
process as follows:—The sense of vision which consists of light rays 
goes out of the eye to the object and, on the way, mixes with the all- 
pervasive light atoms and engenders a beam of light very wide at the 
further end (prthvagram ). This beam, on reaching the lighted spot 
where the object is, mixes with the external light, both being of the 
same kind. This is what is meant by the perfection of the sense of 
vision. The fact that the object is illuminated by the external light is 
its perfection. Thus there takes place perfection of both, according to 
this view.] 

81. According to some, the sound is cognised as 
identical with the word ( sphota ); others think that the 
sound is not perceived at all, which others still think 
that it is perceived as a separate entity. 

Those who believe that the word is manifested hold three 
different views on the subject. The sound which is closely 
bound up with the word is perceived as one with it, as the 
colour of the associated object is perceived as one with the 
crystal. According to others, just as the senses, and their 
qualities, being themselves unperceived, 1 become the cause 
of the cognition of the object, in the same way, the sound, 
its own form remaining unperceived, becomes the cause of 
the cognition of the word. Others still say: from a distance, 
even if the form of the word is not perceived, the bare sound 
is perceived. Others explain that it (the perception of the 
sound) is similar to that of the word. It is like this: it is 
seen that in deserts, small things appear to be big. Moon 
and other objects are perceived as small, trees etc. are vague¬ 
ly perceived but not that they have bark and holes on them 
or that they belong to particular species like Dhava or Khq,- 

84 vakyapadiyam of bhartrhari 

[1. The Saiikhyas consider the senses to be products of Ahahkara, 
but the illustration here is based on the view that the senses are pro¬ 
ducts of the elements. Even if they are products of the combined ele¬ 
ments (pahdkrta ), one of the elements is predominant in the combi¬ 
nation. The predominant element with its special quality, is itself not 
perceived when it becomes the cause of the perception of the same 
quality in the external products. The sense of smell has smell as its 
quality but that is not perceived when it helps us to perceive the smell 
of other objects.] 

82. Just as a Vedic passage ( anuvaka ) or a verse 
is well-fixed in the mind 1 after the (last) repetition and 
is not fully grasped in each repetition, 

The sounds relating to the phonemes, words and sentence, 
produced by special efforts, manifest the stfiota of the pho¬ 
nemes, words and sentence and superimpose them on the 
cognitions. If the ultimate parts of the phonemes 2 are per¬ 
ceived gradually, could be no collection of them and the 
final cognition would be devoid of an object. This point has 
been discussed in many ways in the commentary 3 on the 
Bhasya on the sutra which defines connected text ( samhita ). 
Even when the form of the whole word is manifested, as long 
as it has no definite form and special features, it is as good 
as unperceived and no usage can be made of it. 

[1. Sodhatvam. This is an unusual word. Vr. explains it as be¬ 
coming the objects of one act of remembrance: smrtibuddher ekasyd 
visayabhdvam. It amounts to becoming well fixed in the mind. 

2. Kramena tu vai-Tiaturiyagrahane. What is pointed out here is 
that if it is held that the parts of a sphota are gradually manifested, it 
cannot be manifested at all. Each sound manifests the whole of the 
sphota. Thus it is manifested again and again. 

3. Samhitdsutrabhasyavivarane. The sutra is P. 1.4.109. The 
Vivarana is obviously a commentary on the Bhasya on this sutra, 
written probably by Bhartrhari himself. See p. 9 of Pt. Charu Deva 
Shastri’s Sanskrit Introduction to his edition of Vakyapadlya, Kanda I.] 

83. In the same way, through the previous eogni- 
tions ? unnameable, but favourable to the final cleay 

cognition, the form of the word, manifested by the last 
sound, is perceived. 

The many cognitions, produced by the sounds in the course 
of the manifestation of the word, being conducive to the clear 
perception of the form of the word, inexpressible and in the 
nature of means, become the cause of the perception of the 

The following is the process by which the form of the 
word is clearly perceived: 

84. The word is grasped in the (final) cognition 
the seeds of which have been sown by the sounds in¬ 
cluding the final one and which has gradually attained 
maturity. 1 

The sounds, while they manifest the word, leave im¬ 
pression-seeds 2 progressively clearer and conducive to the 
clear perception (of the word). Then, the final sound brings 
to the mind which has now attained maturity or a certain 
fitness by the awakening of the impressions of the previous 
cognitions, the form of the word as coloured by itself. 3 

[1. Verses 82, 83 and 84 are quoted in the Sphotasiddhi (p. 132), 
(Madras University Sanskrit Series 6). 

2. Vyaktapanccheddnugunasamskdrablidvandbljani. Samskdra, blva- 
vand and bz ja denote the same thing. They stand, according to Vr., 
for three aspects of the same thing. The previous somewhat vague 
cognitions of the sphotri leave their impressions in the mind. Vr. says 
that they are called samskaras because, they, in a way, perfect the 
mind; they are called bhavana because they give them the form of con¬ 
sciousness ( bhdvayanti ) and, finally, they are called seeds ( bljdni ) 
because they are the causes of the later clear cognition. 

3. Upagrahena. Vr. explains this by ‘svikdrenci’. This is probably 
a reference to the fact that the sound colours the sphotti with its 
own form.] 


VAkyApadiyam of bhartrhari 

85. When one perceives those non-existing ele¬ 
ments in the middle as existing, it is due to the incapa¬ 
city of the perceiver, they are only a means to the 
cognition of the real word. 

When the indivisible, sequenceless and unbreakable 
phoneme, word or sentence is manifested by the sounds, 
cognitions appearing to perceive parts of the phoneme and 
relating to its parts occur in regard to the phoneme; cogni¬ 
tions appearing to perceive phonemes as parts of the word 
occur in regard to the word and cognitions appearing to per¬ 
ceive words as parts of sentences occur in regard to the 
sentence. Due to these cognitions, hearers think that these 
non-existing parts actually exist. It is only the incapacity of 
the hearers who can only perceive the word as shown by 
others. 1 Indeed their perception of the word can take place 
only through such means, such sequence and such phonemes 
(which appear as parts of something which is partless. 
(Others, on the other hand) perceive and cause others to 
perceive the sequenceless word, even a dasatwya, 2 through 
other means. 3 

[1. Parapradarsita. What is emphasised here is that ordinary per¬ 
sons can perceive the word (sphota) only through the sounds and as 
possessing the attributes of sounds. The next sentence makes this 
point clear. 

2. Dasatayd. The word occurs also in the Vrtti on I. 51. There, 
as here, Vrsabha explains it as catuhsastih — sixty-four. 

3. Upayantarena Vrsabha explains it is arsena jnanena.'] 

86. The appearance of diversity in knowledge and 
the Word is surely an illusion. The word is coloured by 
sequence and knowledge is coloured by the known. 

Even though knowledge is undifferentiated and formless, 
it appears to be differentiated because it takes on the form 
of all the things which are to be known, as when one says 
five trees, twenty cows. The inner word in which..aU the 

I. 87-88. 


seeds are merged, appears, at the time of its manifestation, 
to follow the differentiation and sequence of the manifesting 
sounds. Therefore, the indivisible word-principle, called 
Word-Mind, being affected by the appearance of difference 
belonging to something else, is understood differently from 
what it is. It has been said: — 

“Knowledge, without the thing to be known, does not 
enter into usage. And nothing can be expressed through the 
word which has not assumed sequence.” 

87. Just as the cognition of the earlier (lower) 

numbers is the means for the understanding of the 
number in question, though they are different from one 
another, in the same way, the understanding of the 
other verbal elements is the means for the understand¬ 
ing of the sentence. ' 

Just as one who wants to cognise the number hundred or 
thousand which differentiate their substrata, understands as 
means thereof the numbers one etc. having different effects 
as parts of the numbers hundred etc. 1 in the same way, the 
cognitions of the different words Devadatta etc. is the means 
for the perception of the form of the sentence. Hence their 
acceptance is inevitable. 

[1. The illustration of the cognition of numbers is based on the 
Vaisesika conception. The Vaisesikas believe that from duality on¬ 
wards up to the highest number p ardrdha, all numbers are produced 
by apeksabuddhi, ie. the notion that relates to many unities before 
the next number is produced. When two things are brought before us, 
we cognise each one separately as ‘this is one and that is one’. This 
is apeksabuddhi. Then arises the notion of two. This is true of all 
the subsequent numbers. Thus the previous numbers become the 
means for the production and cognition of the later numbers.] 

88. Even though the sounds manifesting the 
phonemes, words and sentences are entirely different 
from one another, their powers appear confused. 



Separate efforts are made in regard to the phoneme, word 
and the sentence and the air set in motion by them strikes 
against the places of articulation. Even though the sounds 
thus perfected by contact with the different places of articu¬ 
lation are different from one another, the difference between 
these manifesting agents is difficult to grasp as in the case of 
the manifesting substratum of the universals of the cow and 
the gayal or as in the case of the substratum of the actions 
of turning and pouring. Even though the manifesting powers 
have separate effects, their essence, possessing some com¬ 
mon features, is confused due to some special effect. Hence 
there is the comprehension of parts in the partless phonemes, 
of divisions in the form of phonemes in the word and of 
divisions in the form of words in the sentence! 

[1. The Vrtti explains why the partless phoneme, word and sen¬ 
tence appears to have parts. The sounds which manifest the^ phoneme 
V are different from those which manifest the word gaulj, and the 
latter are different from those which manifest the sentence gamabhyaja 
But they are identified with one another because of their apparent re¬ 
semblance. They differ from one another in their cause as well as in 
their effect. So their powers are different. They suggest such different 
things as phoneme, word and sentence. This is difference in the effect. 
They are produced by different efforts, such as the effort to manifest the 
phoneme, the effort to manifest the word and the effort to manifest the 
sentence. This is difference in the cause. Difference in the effoit causes 
difference in the air which strikes at the different places of articulation. 
Though these manifesting sounds are different from one another, the 
difference is not noticed, because, on the basis of imagined resemblance, 
they are identified with one another. The sounds manifesting g are 
identified with those manifesting V i* 1 9 au h these latter with those 
manifesting { g’ in the sentence c gamdbhyaja \ So the word which is indi¬ 
visible appears to have parts e g 3 etc. and the sentence which is partless 
appears to have parts c gam’ etc. This is what is meant by sanklrna 
iva saktayah the powers seem to be confused. The appearance of parts 
in the partless is the result of this confusion. If, for the manifestation 
of two indivisible words, one has to make similar movements of the 
vocal organs, the phonemes produced by these movements, appear to 
be parts of both of these indivisible words.] 

89. Just as, from a distance or in the dark, after 
having mistaken an object in the previous cognitions, 
one sees it correctly, 

I. 90-91. 


It is in accordance with the properties of objects and the 
senses that people of ordinary vision, perceiving from a dis¬ 
tance, only the general shape, mistake trees etc. for elephants 
etc. Staying in the same place, by steady concentration, gra¬ 
dually, they perceive their special features. Entering sudden¬ 
ly dimly lighted rooms from a well-lighted place and after 
having mistaken rope etc. for snake etc. they, once their 
eyesight has become normal, through steady concentration, 
perceive them as they are. 1 

[1. Cf. Sphotasiddhi, verse 19 and the commentary thereon. It is 
an echo of this verse and the Vrtti thereon. This and the following 
verse are actually quoted there (p. 143.)] 

90. In the same way, when the sentence is being 
manifested by the sounds which are the causes of its 
manifestation, there is at first a cognition in which 
parts figure. 

Even though the sentence is indivisible, the sounds pro¬ 
duced by the efforts intended to manifest it, cause a know¬ 
ledge consisting of the cognition of parts appearing as phone¬ 
mes and words, because of resemblance in their powers which, 
in reality, are different. 

91. Just as there is a definite sequence in the 
transformation of milk or of the seed in the same way, 
there is a definite sequence in which the cognitions of 
perceivers take place in regard to the word (sentence) . 

Just as, according to the creation theory or the manifes¬ 
tation theory, milk, prompted by its transformation clarified 
butter etc., passes through definite stages like curds, having 
or not having their special names, preserves, as far as possi¬ 
ble, the presence of the properties of the original material 
and, without violating the fixed sequence of the stages, shows 
the transformation which prompted it; or rice-seed, prompted 
by its transformation, the rice-grains, passes through the 
inevitable stages like stalk, sprout etc. and appears as its main 
K. S. 13 



transformation which prompted it, in the same way, the cog¬ 
nitions of ordinary hearers, prompted by the main result, 
the understanding of the meaning of the sentence, preceded 
by the grasping of the form of the sentence, show, when the 
effect having its fixed means has to be brought about, appea¬ 
rances of parts having fixed sequence and proceed to perceive 
the sphotcis (the words) 1 . 

[1. The use of the expressions prayuktam, prayojikam, prayojakena 
and prayuktanam is based on the idea, a figurative one, that the effect 
prompts its own cause to act in such a manner that the effect is pro¬ 
duced. Vr. puts it in this way: Pradhdnam Kdryasytitmalabhaya Kara- 
Tiani prayunkte = Primordial matter sets in motion the causes in order 
that the effect may come to be. The activity of Primordial Matter is 
attributed to the effect itself. This verse is quoted in Sphota-siddhi, 
p. 159.] 

92. Even if the words are accepted as having 
parts, the difference in parts is due to the sequence of 
sounds. And if they are partless, the appearance of 
parts (due to the sequence of sounds) is the means for 
their cognition. 

According to those upholders of division who are of the 
opinion that the word ‘gauh’ is nothing more than the 
phonemes g-au-h, that there is no indivisible word-essence 
beyond them to be cognised through the cognition of the 
phonemes and who accept the eternality of the words, when 
the unnameable minute parts in the form of phonemes are 
gradually manifested, the form of the word would not be 
cognised and the final clear cognition of the form would have 
no object to be cognised. If it is held that all the parts are 
manifested at the same time, there would be no difference 
between vega and gave or tena and nate as far as the audible 
side is concerned. If they are looked upon as two divisible 
words with two different meanings, this difficulty would not 
arise. According to that view also, their perception as two 
different words would be due to the difference of sequence of 

I. 93. 


the manifesting sounds. It is seen that ropes etc., illuminated 
by a lamp fed with the fat of frogs etc., appear like snakes etc. 

How, according to the view that the words are indivisible, 
the cognitions embracing different powers resulting from the 
postulation of divisions become the means (of cognising the 
word) has been explained in the immediately preceding 
verses. 1 

[1. Three views in regard to the word are referred to in this verse. 
From ‘Ye’pi to ‘note iti ca’, the view discussed is that the phonemes 
themselves constitute the word and that there is no word beyond them. 
From ‘tatra sabddntdre’ to ‘sarpadipratipattibhedali’, the view mentioned 
is that the word is eternal and so are its parts. From ‘nirbhdgesu’ till 
the end, the view is that the word is indivisible.] 

93. Some have declared the universal manifested 
by the individuals to be the sphota and what are called 
sounds ( dhvanis ) to be the individuals. 

Some, proclaiming the eternality of the word because of 
the eternality of the universals, declare that in statements 
such as the following 

“On both sides, it is only the sphota which is shown: in 
place of the phoneme V comes the phoneme T.” 1 

It is the universal of the word which is called the sphota, 
it is different from the universal called Svord-ness’ and its 
fitness to be perceived is brought about gradually by its 
substrata which are produced in a sequence and do not co¬ 
co-exist. The individual sounds which are produced and 
which make the unnameable sphota nameable get the name of 
sound ( dhvani ). 2 

[1. M. Bha I. p. 26, 1. 1. 

2. Before, in verse 69, it was declared that the universal is the 
name and that the individual is the named. Here, the view that tba 
universal is the sphota is set forth.] 



94. The sound, modified by its causes, becomes 
the cause of the cognition of the changeless word, just 
as the light from a lamp (becomes the cause of the 
perception of an object). 

Others, on the other hand, on account of difference from 
the process which results in the use of the terms universal 
and particular, assert that there is only one word which is 
external, not subject to any change through the sounds modi¬ 
fied by their own causes and not residing in the sounds. It 
is manifested as is done by the light of a lamp when there 
is no vagueness in the object, with the colouring of the 
changes in the manifesting sound, as though it has received 
a new form. 1 

[1. Another view of sphotd is referred to here. It is not clear who 
held this view. Vr. gives the following quotation expressing this view: 
Tasmin nitydrn, sabdasaktivi pratijdnate — they assert that there is 
one eternal word-power in it (the ether). The main idea here is that 
the sounds, which differ from one another because of difference in the 
articulatory movements, cause the cognition of the one changeless word 
without effecting any change in it, just as the light from a lamp reveals 
the object without effecting any change in it. This process is said to 
be different from the one relating to the universal and the individual in 
wholes like a jar: akrtivyaktivyavaluLravaidharmydt. The parts of a jar 
are the substrata of the whole, namely, the jar. The jar itself, which is 
the individual, is the cause of the cognition of the universal in it and 
of the application of the word ‘jar v to it. This is not the relation bet¬ 
ween the sounds and the word, because the sounds cannot co-exist. As 
Vr. puts it: Yaugapadydblvdvdc chabdabhdgdndm .3 

95. Manifestation is not necessasily confined to 
that which is transitory. The manifestation of the 
eternal universals by their substrata is admitted. 1 

Some put forward the fact of manifestation itself as a 
reason for concluding that it is non-eternal, as follows—The 
word is non-eternal, because it is manifested, like a jar. It 
seen that non-eternal objects like a jar are manifested by 


I. 96. 

a lamp etc. The word is manifested by sounds; therefore, 
it is non-eternal. If it is held that it is not manifested, that 
would mean that it is produced. That also makes it non¬ 
eternal. Now, to those who hold that universals exist and 
that they are eternal, it is answered; the reason given (in 
the above syllogism) is not a real reason, because it is seen 
both ways. Against those who hold nothing to be eternal and, 
therefore, maintain that the invalidity of the reason is not 
proved, those who consider the word to be eternal declare 
even before them that the invalidity is not established. How? 
Against the opponent (one who does believe in eternality) 
the reason “because it is manifested” is not a good reason 
because it is doubtful and would lead to option and as the 
other incidental properties also would be attributed, the dis¬ 
cussion would lead to regressus ad infinitum. 2 

[1. Verses 95 to 101 answer objections to the manifestation theory. 
The objections are made clear in the Vrtti. 

2. Prasahgo vanusanginam ityanavastha syattarkcisya. If the word 
is said to be non-eternal, because it is manifested, like a jar by the 
light of a lamp, why not attribute the other properties of the jar, like 
visibility also to the word? There would then be no end to the process 
( anavastha )]. 

96. It is only material objects which are associat¬ 
ed with particular places. Even when difference of 
place between sound and word is imagined, there is, 
in fact, no real difference. 

Another opposite view is now put forward: The word 
is not manifested, because of difference in location between 
the manifestor and the manifested). Jars and other objects 
are manifested by lamp etc. because they exist in the same 
locality. The word is found in a place other than where its 
manifestos, the contacts and separations of the vocal organs 
take piace. 1 This objection does not really hold good in the 
case of the sounds, Still, it is argued as follows—How is 
the word which is in one place manifested by the sounds 


which are in different places and very far from one another? 
This is answered as follows—To be in one place or to be 
in many places is a property of corporeal things. Even 
material objects like the sun, while being actually restricted 
to one place, are perceived as being in many places. Sound 
and word, both, incorporeal, are beyond the range of such 
expressions as location and located and, therefore, even 
though the misconception of difference of location takes 
place, really speaking, there is no difference between them 
in this respect. 

[1. The opponent s idea is that the manifesting sounds exist in the 
vocal organs whose contacts and separations produce them, whereas the 
word exists in ether (akasa).] 

97. Just as there is an eternal fitness between the 
senses and the objects, in the same way, there is the 
relation of manifestor and manifested between parti¬ 
cular sounds and the word (sp hota ). 

Another opposite view is now put forward:. The word is 
not manifested because the (socalled) manifestators are fixed. 
In this world, what is to be manifested does not require a 
fixed manifestor, since all objects like jars are manifested 
by any one of the following, a precious stone or a lamp or 
a luminous plant or a planet or a star. In the case of words, 
on the other hand, it is held that fixed sounds manifest them; 
the sounds which are the causes of the manifestation of parti¬ 
cular phonemes cannot manifest other phonemes. Therefore, 
words are not manifested. To this, the verse which begins 
with the words: “between the senses and the objects (grahana- 
grahyayoh) is the answer. Just as the colour which inheres 
in the sense of vision is the cause of the manifestation of 
external colour, not other qualities nor other senses, nor the 
qualities of other senses, in the same way, they (the sounds) 
become the causes of the manifestation of external objects. 1 

[1. The answer to the present objection consists in pointing out that 
even in regard to other manifestos, like the senses, there is a certain 

I. 98-99. 


fixity and restriction. While the sense of vision, which is of the nature 
of fire according to the Vaisesikas, can reveal the colour of external 
objects, it cannot reveal their smell or taste.] 

Against the argument that such fixity does not exist 
where a thing is manifested by a sense having the same 
attribute as itself, the following answer is given— 

98. In the case of smells etc., which are manifest¬ 
ed by their similars, it is seen in the world that there 
is a particular factor in each case which is responsible 
for it. 

Even when things are perceived by senses having the 
same attributes as themselves, this fixity can be observed. 
For example, among substances like nail, bitumen etc., only 
some particular substance in contact can manifest the smell 
of some particular substance. 1 

[1. When the colour in the sense of vision manifests the jar, it is a 
case of difference between the manifestor and the manifested, because 
the jar is a substance and the colour in the sense of vision is a quality. 
Where there is such difference, the opponent points out, there is no 
requirement of a fixed manifestor. But where there is similarity bet¬ 
ween the two, there should be fixity. But in the case of the word being 
manifested by the sounds, there is similarity, because both are cognised 
by the sense of hearing and yet there is a fixed manifestor. Against this, 
it is pointed out that, in other cases also where there is similarity, a 
fixed manifestor is required. The smell of an object is perceived only 
by the sense of smell. Thus the manifestor and the manifested are simi¬ 
lar to each other. And yet smells are manifested only in contact with 
a particular substance. The smell of Kunkuma is manifested only in 
contact v/ith clarified butter. The only when it comes in contact with 
boiling barly broth.] 

99. The thing that is revealed follows the differen¬ 
ces of the revealors. This is evident in the case of 
reflections in such different reflectors as oil, water etc. 

Another opposite view is now put forward—The word is 
j§ not manifested because (in its case) increase, decrease 



and difference in number of the manifestors are seen in the 
manifested (the word). When there is increase and decrease 
of the manifestors, increase and decrease of the manifested 
are not found. Increase and decrease of (the light of) lamps 
(the manifestors) do not cause increase and decrease of jars 
etc. Nor does any variation in the number of lamps cause 
any change in the original number of jars etc. In the case of 
the word, on the other hand, one sees variation in number 
and magnitude following variation in the striking of the vocal 
organs. Therefore, the word is not manifested. 

This argument is answered as follows—one does see that 
the manifested follow the variations of the manifestors. It 
is as follows—When the surface of the mirror is sunk, the 
reflection of the face is raised, when the surface of the mirror 
is raised, the reflection of the face is sunk. In a dagger, the 
reflection of the face is long, in mustard oil, it is dark, in a 
Chinese dagger, Greek crystal etc. it follows the dimensions 
of the reflector. Thus possibilities of variation are infinite. 
Difference of number is also seen in the case of the reflec¬ 
tions of the sun etc. when there are many mirrors and many 
waves of water. 2 

[1. It is not clear what specific objects are meant by cmasastra and 
yavanakaca. Obviously, they could reflect objects. 

2. The Vrtti may be compared with the following passage from the 
Mahabhasyatika of Bhartrhari— 

“Yasyapi sabdavyaktih tasynpi nityah sabdali. Sa tu nadabhivyang- 
yah. padaniyato nadah. Yathd caksurddayo niyata abhivyanjaka abhi- 
vyangyesu rupddisu. rupavrddhirdsdnuvidhdyinasca . Yathadarsaman - 
daladisu pratibimbdni dirghani parimandaldni mahanti anydni ca dri- 
yante evam sabdd api nddabhedena bliidyante. Yatha saVde tarahgabhede 
naikas candro 9 neka upalaghyate, pradApabheddcca chayd bhidyate 
adarsabheddcca pratibimbabhedah. Tasmdnniyatdndddbhivyahgyji hrada - 
vrddhihrdsdnuvidlidyino vyaktisabdd api nitydh . 

Mahabhasyadipika of Bhartrhari, p. 20 (A.B.O.R.I, XLIII). 

It might be said that the reflection which has entered a 
nnrror etc. is different from its originals like the moon etc. 
Against this, the answer is—• 

I. 100-101-102. 


100. Original objects like mountains cannot 
possibly enter into reflectors like a diamond and the 
surface of a mirror which have totally different dimen¬ 

It is not possible that objects like mountains having diff¬ 
erent dimensions should enter and be produced inside sub¬ 
strata like a diamond. 

101. Therefore, due to the difference in time of 
the manifesting sounds, their own time and that of the 
duration of their cognition are attributed to the 
phoneme (varna) word (pack) and sentence ( vakya ) 
whcih have no difference in time. 

In the continued existence of eternal things, the power 
of time as an auxiliary cause does not play a part. All the 
sphotas, the phonemes, words and sentences, having a greater 
or lesser magnitude and whose continued existence, during 
the, according to wordly usage, earlier and later periods, is 
understood from their cognitions (at the time of their mani¬ 
festation) do not differ as far as their duration is concerned. 
When they become objects of cognition, though they have no 
difference in time, they seem to have the duration of their 
cognition. What is meant by speed and their own duration is 
this—the primary sound whose duration is superimposed on 
the essence of the word, in order that their difference may 
not be grasped, becomes the cause of the use of the expression 
short, long and protracted. The secondary sound brings 
about the external difference of speed, quick etc. 

102. Others have declared that whatever is pro¬ 
duced by the organs of articulation, through contacts 
and separations, is the sphota; the sounds produced by 
this initial sound are the dhvanis. 

K. s. J? 



According to the view that the word is transitory, the 
first sound which is produced by the contacts and separations 
between the points and organs of articulation is the sphota. 
Those which are produced by that first sound and spread in 
all directions carrying its reflection, are the dhvanis. All 
things are in themselves without parts, but the parts of their 
main associates are attributed to them. Similarly, ether 
(which is one) has no parts but division into parts due to 
relation with objects in contact with it, is attributed to it. 
That being so, because of proximity in the form of succession 
of parts and on account of uninterrupted succession of cause 
and effect, those elements which have progressively diminish¬ 
ing power of reflecting the previous sounds are similar to the 
forms illuminated by dim light, gradually disappear and cause 
division in the phoneme, are called sounds. According to the 
view of eternality, the sphota is manifested by sounds 
produced by contacts and separations. According to some, 
it is manifested by the resonance 1 produced by the 
sounds caused by contacts and separations. Sounds, on the 
other hand, are those which favour the cognition of the sphota 9 
have progressively diminishing power of manifestation, cause 
the distinction of the speeds quick etc. and gradually disappear. 

[1. Dhvani and ruada. Here also, as in the Vrtti on verse 47, a dis¬ 
tinction is made between dhvani and ruada. The latter is produced by 
the former. This distinction is mentioned in the course of the exposition 
of the views of those who held the sphota to be the one indivisible word 
( sabdavyakti ) as distinct from those who held it to be the Universal. 
According to Vrsabha, the sounds produced by contacts and separations 
produce a resonance (uada) which manifests the sphota. The sounds 
produced by the first sound are the cause of the continued cognition of 
the sphota or of the speed of utterance which is only another way of 
looking at the same thing. See note 1 on verse 47.] 

103. Whether the sound in question is short or 
long, the time of the sphota is invariable. The series 
of sounds which follows is susceptible of greater or 
lesser duration. 1 

r. 104. 


basis of the common feat “* attnbuted f to the the 

amount of space. Or S* T C ,° nS1St 7f 1 of coverin - a ce rtain 
usage is the basis f Everywhere 

Determining the nature o{ ft* ° f ° bjects - 

tradition, would be unrel" ° D ^ baSlS ° f reaSon and 

different views. As botb^ 6 bec ^ use W0lM result ^ 
those which are cause "7 ** which are effects and 

distinction, there i s not the perisb wltbout 

big and a small sound i sllghtest dlfference between a 

of an elephant an a ’ tbere 1S none between th e cognitions 

ot an phant and a mosquito. Due to difference in cir¬ 
cumstances, however th« r, r , , 1 C11 

differs. The succession P °f « “7 *° Pr °? UCe effecls 

one produced by the imoarf T ,, ° , a soun *be 

, * . “f, e “npact of the drum and the stick 

reaches far. Another sound like the one produced by strikin- 

he hlard onlv f"“ “““ 3 TOi °" °* ^ds Jhich can 
be heard only from near, but reverberates without a break. 

the'mmSteta'am ,j ™ h “‘ ^2* '" '™!” * he of 

j j u , , ^ * a accorc iing to the view that the first sound 

Z 7. „ “ Pata,1 ” ! is «“ «**■ and the later “Zl 

produced by the first one are the dhvanis. 

2. If one looks upon sound as a quality, one can object to its 
being: described as ‘small’ or ‘big’ because smallness or b gness am 
quabties and they can reside only in substances and not in £ so£d 
which is itself a quality. If 0 ne looks upon sound as a substance, 
even then one can object that only material substances can be big or 
small and not immaterial substances like sound. To meet this objec¬ 
tion, it is pointed out that these two expressions are used for sound 
by courtesy (upacaryate ).] 

104. From a distance only the sound is perceived 
like the light from a lamp. The difference is clearly 
visible in the sounds which are produced by hells, etc. 

Here some thinkers consider that the sphota manifested 
by the sound which comes into being at the same time and 
spreads far in all directions like the light of a lamp is asso- 
ciated with the sound from the very moment of its manifesto- 



tion just as a substance is endowed with its smell from the 
very moment it is produced. Just as, in a lamp, there is 
the burning substance which is the material cause of its closely 
packed parts (ghanasannivistavayavam pratyupaddnam) 
and its light is based on it and follows its transformations, in 
the same way, the sphota and the sound, very distinct when 
a bell is struck, constitute the characteristic of the manifesta¬ 
tion of the phonemes. The difference between primary and 
secondary sounds has been explained in the two preceding 
verses. 1 

[1. This verse puts forward another view of those who held the 
sphota to be transitory. It is an answer to a possible objection that 
the sound, apart from the sphota , does not exist at all. It is here 
pointed out that it does exist separately. We can see the difference 
when we hear from a distance the sounds which a crowd makes but 
cannot distinguish the words (the spliotas). Or, it is like seeing the 
light of a lamp without seeing the lamp itself. The sound produced 
by the first impact of a bell is the sphota. In this verse, it is stated 
that the sound ( dhvani ) is also produced at the same time whereas 
in the previous verse it was stated that the sound produced by the 
impact is the sphota and the subsequent sounds produced by the first 
one are the dhvanis. In both the views, the sphota is transitory. 

In the text of the Vrtti, ghatasannivistdvayavam is a mistake for 
ghanasannivistavayavam .] 

105. On account of the relative intensity in the 
contact between the organs and the places of articula¬ 
tion, the long and the protracted also become different. 
The sounds produced after the cessation of the vibra¬ 
tions bring about variation in the speed of utterance. 

Even according to the view that the word is transitory, 1 
it is not because of any increase in the manifesting sounds 
that there is any increase in the long and protracted vowels. 
How then ? Because of the greater intensity in the contact, 
between the places and the organs of articulation. All that 
contact can only bring about the form of the long and pro- 

i. 108-107. 


tracted vowels. Therefore, till they acquire their form, the 
vibration, causing mutual contact of the parts of the places 
and organs of ai'ticulation because of their striking against 
one another and following closely the special movements of 
the air, continues. When the vibration ceases, the sounds 
which are produced by the original ones, become the cause 
of the regulation of speed like quick etc. 

[1. I have adopted the reading ‘anityapakse’ as Vrsabha has it.] 

106. Even after the organs have ceased to vibrate, 
other sounds are produced from the sphota as one flame 
from another. 

Even when the vibrations continue uninterruptedly, 
whatever sound is produced by the contact (between the 
places and organs of articulation) never disappears without 
producing its effect. The sounds which are produced by the 
vibration and are simultaneous (with the sphota) manifest 
the sphota. The sounds which come in between, resemble them, 
and shine like them, are called secondary sounds (anusahga) 
It has also been said that each secondary sound among them 
has its series of effects lasting as long as the vibration lasts 
and helps in the manifestation of the sphota. This production 
of sounds has been described on the analogy of the continuity 
of the flames proceeding from burning fuel. Just as from the 
flames of burning fuel proceed continuously other flames pro¬ 
duced by them and manifest objects by their light, so is the 
continuity of the sounds. 

107. Air, atoms, or knowledge is said to be trans¬ 
formed into the word according to some. Thus, in the 
different doctrines, there is no fixed view on the 

Some declare that it is air which becomes the word. 


VAkyApadiyam of bhartrHaM: 

108. The air set in motion by the effort corres¬ 
ponding to the desire of the speaker, strikes at the 
different places of articulation and is transformed into 

109. By the impact caused by the force of the 
original cause (the effort) even the solid forms of air 
which has speed and solidity are diversified. 1 

This and similar views must be understood. 

[1. Vrsabha understands that the air brings about the contacts and 
separations of the places and organs of articulation which are solid 
(saravatyah). The translation connects ‘saravatyo ’pi murttayah’ with 
‘tasya’ = of the air.] 

Others have accepted that it is the atoms which become 
the word. They have declared as follows— 

110. Because the different kinds of atoms have 
all powers, they, in combination or otherwise, trans¬ 
form themselves into shadow, sunlight, darkness and 

111. The atoms called words, when their power is 
manifested by effort, are set in motion by the same 
effort and they gather like clouds. 

All such views have to be understood. Others expound 
the transformation of knowledge into the word. 

112. This inner knower, at first identical with 
the subtle word, transforms himself into the gross 
word, in order to manifest his own form. 

t- 113-114-115. 


113. He, the inner knower, becomes mind after at- 
taining matunty through heat. Then he enters the air 
called breath which is then emitted. 1 

7St d thf b T f ? e St6PS by Which the self Hornes the 
r/ L S o beC ° meS the capable of cognising 

°bj' ects . , T . .. c fP acit y attains maturity through the bodily heaf 

men se . in T 10 u- by . the desire to s P eak > the mind identifies itself 
with the breath which is then emitted.] 

114. Air becomes the substratum of the mind and 
becomes coloured by the qualities of the latter and is 
transformed by the inner heat. 1 

[1. The Self becomes the mind of which the air becomes the 
substratum. The air becomes impregnated with the attributes of the 
mind and becomes the word.] 

115. The inner air (prana), after splitting into 
parts in the form of sounds and after suggesting the 
phonemes, is merged into them. 

All this and similar views have to be considered. What 
has been given is only an illustration. There is much diver¬ 
sity in the views of the authors of the Siksas and the 
Bbasyas! For example— 

“Vital air, prompted upwards by the inner effort, aided 
by the inner heat, accumulates, through the sound-carrying 
.interstices, 2 fine particles of sound like masses of smoke. 
When thus accumulated in the places of articulation, it 
assumes the identical form of the inner word, because of its 
illuminating aspect”. All such views have to be considered. 
\t is as follows 3 — 

“The air, prompted by effort from the region of the navel, 
going upwards, strikes against one of the places like the chest 
and then sound is produced”. 

Such diversity of views of the authors of the Siksas 
must be considered. It has been said by the teacher (Panini). 



“The Self sees things through the intellect and joins the 
mind with the desire to speak. The mind strikes at the heat 
of the body which, in its turn, prompts the vital air.” 4 

And so on. Another teacher says— 

“The vital air reaches the stomach where the resonance 
takes place. Reaching the throat, it becomes either breath 
or sonority”. 

And so on. Another teacher says— 

“The heat of the body, struck by the mind, prompts 
the vital air. It rises from the navel and strikes against the 
head and when it comes into contact with another rising 
wave of air, becomes sounds such as c k’ or e kh’” 

Such different views are found in the treatises called 
Siksa attached to each branch of the Veda and they must be 
understood in detail. 

[1. Vrsabha says that by Bhasyakara, the commentators of the 
Siksas are meant. 

2. Vr. says that the veins (wadis) are filled with fine parts of the 
word and when they are set in motion along the veins by the inner 
air, they accumulate in the places of articulation, like masses of smoke 

3. Vr says —Tathety apis ally asikslidarsanam. 

4. Paniniyasiksa, 6.] 

116. The eternal word which, being fine, is not 
perceived by the ear, becomes perceptible through its 
own cause, as the air by fanning. 

Now another tradition 1 is being recorded. There is sub¬ 
tle sound within and without all embodied objects like masses 
of fine air. According to some, it is understood as ether 
Just as, even though atoms of air exist everywhere it •' 
only when, due to the impact of the fan ,2 they are dislodged 
rom their place, that they become endowed with action^ in 
toe same way, sound, when given a gross modified form’ by 

I. 117-118. 


its manifesting causes, reaches the region of the ear and 
perfects it. 

[1. According to .Vrsabha, this tradition is also current among the 
authors of the 6iksas. 

2. Both in the Karika and in the Vrtti, Pt. Charu Deva Shastri’s 
edition has Vyanjana. But the context requires Vyajana and it is actu¬ 
ally found in the manuscript ‘S’, used for our edition of Vakyapadiya 
Kanda I. So that reading has been adopted.] 

117. The power of it which is in the breath and 
in the understanding strikes at the different places of 
articulation and becomes differentiated. 

All these are different views. The sound mentioned in 
the previous verse having the attribute of accumulation is not 
refeired to here. The word, already under discussion, is 
being explained according to different views. The word rests 
on the bieath as well as on the mind.* Being manifested 
by the powers of the two substrata, the breath and the mind, 
the word conveys the meaning. The breath is penetrated by 
the mind. Surging upwards like a flame, following the effort 
which set it in motion, it strikes against the points of articu¬ 
lation of the phonemes and assumes a form favourable to the 
comprehension of the eternal word. Having thus attained the 
state of the effect of the two powers which are within it, 
the breath assumes different forms, as the earth or the foetus 
or the banyan seed does 2 and gives to the one essence of the 
word the mere colouring of difference. 

[1. The word has two substrata, the mind and the breath. While 
explaining this idea, Vrsabha adds a third one: the vocal organs. As 
he puts it —Sa cayam trtiyasabdah Karanadhisthanah. 

2. The earth, the foetus and the banyan seed go through certain 
stages before they attain their final form. Unfortunately, the text of 
the Paddhati here is corrupt, so that one does not get a clear idea of 
what Vrsabha considered to be the stages.] 

118. The power which creates and regulates this 
universe rests on words, It is through that eye that 

K. S, 14 



all this diversity of understanding (bhedarupah prati - 
bhatma ) is perceived. 

According to some, the universals (of words and mean¬ 
ings) rest on the substratum of the subtle word. They 
become manifested when the substratum evolves and appear 
as the expressed meanings and the expressive words. It is 
like what some others say—“All objects merge into the senses 
(as potentialities) and all the senses merge into the intellect 
as potentialities and the intellect merges into the sequenceless 
word as a potentiality. All this activity which goes on 
during dreams and the wakeful state and involves distinction 
into different individuals always exists in the subtle word, the 
supreme cause (in a potential state). Others have also said 
as follows— 

“It is the word which sees the object, it is the 
word which speaks, it is the word which reveals the object 
which was lying hidden, it is on the word that this multiple 
world rests and it is this very word which enjoys after the 

119. The difference between ‘sadja’ and others is 
grasped when conveyed through words; therefore, all 
kinds of meanings depend upon the powers of words. 

Everything depends upon the word which causes its 
cognition and enters into usage when grasped through re¬ 
membrance of it as intertwined with its word ( smrtinirii - 
pand) , as identified with its word through memory (abhijol- 
panirupand ) and as connected with some particular action 
(akaranirupan'tt) . The distinction between sadja, rsabha 
gindhdra, dhaivata, nisi da, pancama and madhyama, the 
words expressive of which are not well-defined nor well- 
known, cannot be understood without knowing the words on 
which depend their understanding. Cowherds, shepherds and 
others invent special words and accomplish their special 
purposes in regard to cows and other animals. Therefore, 
a thing with its distinctive feature, closely linked with general 

I. 120. 


or special words the meanings of which are clearly grasped 
or otherwise respectively, is illuminated, embraced and iden¬ 
tified with a cognition which is united with the power of the 
word, intertwined with the word and has the form of the 
word . 1 

[TJhe purpose of this verse is to show that distinction between things 
can be known only through words. The distinctive features of some 
things in this world are easy to see and they have their words to express 
them. But the distinctive features of some things are not easy to grasp. 

By merely listening to the musical notes sadja etc., one cannot under¬ 
stand their special characteristics. One can do so only with the help 
of the technical explanation of the Science of Music. What we cognise 
can enter into worldly usage only if we can express it in words. The 
process of cognition has three stages which, as explained by Vrsabha, 
aie * (1 to see a thing and to remember it as intertwined with its 
word (Smrtinirupayjd), (2) to cognise the identity between word and 
meaning^ (abhijalpanirupana) , (3) to see it as connected with some 
action ( dkaranirupcnia ,). 

The intimate relation in which the word and the meaning figure 
in t e cognition is emphasised in the Vrtti by the use of several ex- 
piessions coveiing different aspects of it: bhedavdnartliah sabdasakti - 
savisi stay a,~ sabdctnuviddhayd sabd&tviikayd buddhya prakasyate upa- 
9} lyate sviki iycite. The meaning is illuminated and embraced by the 
cognition, it becomes one with it. The word, on the other hand, is also 
intertwined with the cognition which has the form of the word. The 
cognition has the form of the word which has the form of the mean¬ 
ing. (upgrhitarthakarasabdarapd sd buddhih—Vi\)] 

120. Knowers of tradition (the Vedas) have de- i m ? 
claied that all this is the transformation of the word. 

It is fiom the chandas that this universe has evolved. 

, aS °^ er thinkers , 1 while explaining causality, saw 

a l e properties of the cause continue in the effects and 
ave eclared as the source of everything, either the mass 
oi atoms, free from mutual distinctions, subtle, imperceptible, 
having the potentialities of all effects, with the tendency to 
pro uce them kept in abeyance or Primordial Matter or the 
co .ection oi Powers rooted in Nescience or something which 
as no biith nor change (but is merely the substratum) of 
appeal ances, in the same way, in the Scripture also, the word 



iii which the powers of Enjoyer and Enjoyed are submerged 
has been declared to be the cause of the world in many ways. 
For example, it has been said: “The Lord Vairaja, is indeed 
made up of Rk, Yajus and Saman. The Lord is the world, 
the Lord is the sacrifice. In it, three oblations are offered, 
pleasing to the three worlds. These oblations offered in a 
three-fold manner, are the three worlds.” 2 Similarly—“He 
who created the world is indeed the Lord of the world, made 
up of hymns, existing at the very beginning, indestructible 
(or a very bull). From him were born cattle, from cattle all 
vegetation and from vegetation, fire. That is why it has been 
said that one should not milk in a wooden vessel. A wooden 
vessel is, indeed, fire. That is why milking is never done in 
a wooden vessel.” 2 

There is a RK also on this very subject— 

“From the Lord first came knowledge, the food; from it 
were separated name and form; the name came from life 
(breath) and the form from knowledge. The one knowledge 
appears as many.” 3 

Again, it has been said—“It is the word which became 
the worlds; the word became all that is immortal and mortal. 
It is the word which enjoys, which speaks in many ways. 
There is nothing beyond the word.” 3 

There are also some ancient sayings on the subject— 

“The Creator, mentioned in the Scripture, after dividing 
Himself in many ways, into manifestations of Himself, entered 
into Himself with all the manifestations.” 4 

“Those persons in whom the pure speech is established 
in a great measure, in them the holy light of the Creator 
exists eminently.” 

“The great Light of the Creator which is in the learned 
as though covered with a lid, becomes merged with its 
source, when their body dissolves.” 

“When one is in possession of right knowledge, one is 
identified with the shining mass of that wonderful light and 
is merged in it.” 

t i2i. 


tl- The problem here is to decide what anye means. According 
to Vrsabha, it means the Vaisesikas and others. If he is right, then 
the Vrtti “yatliaivdnye ... vyavasthapayanti’ mentions different things 
beginning with the mass of atoms ( anugrdma ) as the original cause of 
the world, according to others. For Vaisesikas, it is the mass of atoms, 
foi Sankhyas, it is Primordial Matter (PrcLdhdncL) , for others, it is the 
bundle of Powers, rooted in Nescience and for others still, it is some¬ 
thing devoid of birth and change and is the substratum of appearances 
(vivarta). The other way of interpreting the passage is to take it 
as describing only the mass of atoms, linking all the adjectives in 

the passage with anugrdmam \ The former way is more in accordance 
with our text. 

2. The source of these two prose texts is unknown. In the second 
text, the word ‘ aksan’ may be ‘uksan. 

3. These two verses are given as RK , but they are not found 
in the Rg-Veda. 

4. The idea in this verse is that all differentiation merges into the 
ultimate at the time of pralaya. That is, it exists in a latent stage, 
only to become patent at the time of creation. As Vr. puts it. vyakti - 
rupena bhutvti saktirapeiiavatisthate.'} 

121. All knowledge of what is to be done in this ' • - 

world depends upon the word. Even the child, with 
its residual traces from the previous birth, has such 

Even that which exists is as good as non-existent as long 
as it does not come within the range of verbal usage. Even 
a totally non-existent thing like a hare’s horn or something 
which appears and disappears in the sky like a celestial town 1 
(gandharvanagara ), when brought to the mind by words, 
figures, like something endowed with primary reality, 2 in 
various usages. 'In children in whom the germs of the word 
exist according to their kind, because of the existence in 
them of the residual traces of their use of words in their 
former births, there arises cognition based on vague words 
(anakhyeyasctbda) in the course of their various purposeful 


[1. Vrsabha points out that a celestial city may suddenly appear 
to us in the sky and it may disappear equally suddenly. But when the 
word brings the idea of it to our mind, it stays. 

2. Mukhyascittayuktam iva. Mukhyasattd (primary reality), the 
fact of something existing outside our mind is contrasted with Upa- 
carasattd which consists in the fact of something figuring in our mind 
through the agency of words. Words move chiefly in the realm of 
Upacarasatta , as explained in Vak. Ill, Sambandhasanuddesa.] 

122. The first movement of the vocal organs, 1 the 
emitting of the air upwards and the striking of the 
places of articulation are not possible without the resi¬ 
dual traces of the speech (in the previous birth). 

This residual trace of speech has no beginning and it 
exists in every one as a seed in the mind. It is not possible 
that it should be the result of the effort of any person. Move¬ 
ments of the articulatory organs by children are not due 
to instruction by others but are known through intuition. 
Who, indeed, can make or make known these human proper¬ 
ties as other than residual traces of the word? 

[1. Vr takes the word ‘Karana’ as denoting the vocal organs as 
well as the mind (antahkarana) . Thus, it is due to sabdabhavarua. that 
t e child makes, untaught, not only the first movements of the vocal 
organs, but the very effort to say something. The first movement of 
the vocal organs may not be necessarily to speak, but to eat or to 

_ _ 2 : Pratibhagamyah. Pratibhd seems to be another name for sab- 
dabhavana. It stands (1) for the residual traces of the exercise of the 
3T 1 ^ ° f u s Pf ch ln th e previous birth; (2) for the faculty of speech 
wrth which the child is bom; (3) for the child’s instinct to do some¬ 
thing in response to a situation.] 

123. There is no cognition in the world in which 
the word does not figure. All knowledge is, as it were 
intertwined with the word 

I. 124. 


When, in the word-seed, everything is merged, then no 
verbal usage can be accomplished with the indeterminate 
knowledge which takes place in regard to objects. For in¬ 
stance, when one walks quickly and treads on grass and 
clods of earth, even though knowledge of them takes place 
(no usage is accomplished thereby). That stage of knowledge 
is only sometimes reached in which, when the word-seed is 
awakened and the powers, confined to particular meanings, of 
expressive, explainable and unexplainable words are mani¬ 
fested, the object is given a shape and accepted by knowledge 
which is intertwined with the word and follows the power 
of the word and it comes within the range of clear cognition 
and can be designated as such and such and is said to be 
known. Where, due to circumstances, the manifesting causes 
of the word appear, it becomes a cause of memory. Similarly 
according to some, even one who is asleep has a stream of 
cognitions like one who is awake. The only difference is 
that the seeds of the word function in a very subtle manner 
in that condition. That is why that condition has been called 
one of darkness. Thus consciousness, mixed up with the 
word, appears and disappears constantly as the cause and 
effect of the manifested word. 

124. If this eternal identity of knowledge and 
the word were to disappear, knowledge would cease to 
be knowledge; it is this identity which makes identifi¬ 
cation possible. 

Just as illumination is the nature of fire or consciousness 
the nature of the Inner Controller, in the same way, all 
knowledge is intertwined with the word. Even in the state 
of unconsciousness (sleep), there is the persistence of the 
association with the subtle word. Also, that first cognition of 
external objects which does not grasp their special features 
( nimitt&nam ) illuminates them in a vague manner ( avya- 
padesyayd vrttya) as mere things, by referring to them as 
this or that. At the time of remembrance also, when the 
seeds of such indeterminate cognition are awakened, a mere 



'outline, consisting of the previous vague cognition, figures in 
the mind in the form: 'this is some hymn or verse which I 
have heard before’. If knowledge were not mixed up with 
the word, the (vague) cognition which arises, not being 
intertwined with the form of another (the word) does not 
become an auxiliary in the act of illumination. After the 
cognition of the bare meaning of the words (in a sentence) 
which are different from one another, mutually unhelpful 
and independent of one another (atmdntar&natmandm) has 
taken place, what takes place afterwards, namely, the mutual 
determination of their meanings, the definite identification of 
their meanings as such and such, their cooperation to fulfil 
one purpose, the connection of the individual sentence-mean¬ 
ing with the expressive power of the words, all this is closely 
linked with the fact of knowledge being closely intertwined 
with the word. It is this (vdg'rupatd) which brings about 
the identification and the mutual determination of the word- 
meanings, a cognition (of the sentence) which is qualified 
by all the qualifications and causes purposeful activity and 
yet it does not abandon the appearance of difference within 
it by a process of abstraction of powers. 2 

[1. Atmantaratmanam. Vrsabha’s text seems to have been: ‘atman- 
tardnatmanam 9 because he explains as follows— dtmdntaram esdm 
dtmti na bhavati . Itaretarasya iti. Itarasyetara atmd na bhavati. Itaras- 
ydpitdro devadattartho gavarthasyatmetyddi. 

2. From f bhinnarupdimm to na vijahdti , it is explained that the 
understanding of the sentence-meaning from the word-meaning pre¬ 
supposes that knowledge is of the nature of the word. The stages in 
the process, as understood by Vrsabha, are: (1) the understanding of 
the word-meanings from the words; (2) the identification of the mean¬ 
ings in a cognition intertwined with the word; (3) the unification of 
these meanings by their mutual delimitation.] 

125. It is this which is the basis of all the scien¬ 
ces, crafts and arts. Whatever is created due to this 
can be analysed (and communicated). 

Human transactions in regard to worldly things and in 
regard to things taught in the Vedas takes place through the 

I. 126. 


sciences crafts and arts. The activity of the rest of the 
animate and inanimate creation depends upon man. All 
science is closely bound up with the intellect (thought) which 
is of the nature of the word. It is on the basis of this fact 
of knowledge being of the nature of the word that in cases 
such as the making of a jar, all instruction and effort of the 
prompter and the prompted respectively, proceed. 

126. The consciousness of all beings going 
through transmigration is in the nature of the word; it 
exists within and without. The consciousness of all 
types of beings does not go beyond this essence. 

It is because consciousness is of the nature of the word 
that the distinction between sentient and insentient is made 
in the world . 1 It has been said— 

It is the word which urges all beings towards purposeful 
activity. If that were absent, everything would be insentient 
like a piece of wood or a wall.” 

The experience of pleasure and pain in the case of 
those whose consciousness is turned inward is possible 
only as long as consciousness is of the nature of the 
word. As for those whose consciousness is external, their 
transactions in the world are dependent upon that and 
would, in the absence of that, cease to be. There is no 
kind of being endowed with consciousness who knows 
or causes others to know without the association of the word . 2 
Therefore, there is no activity of consciousness which is not 
closely linked with the powers of the word. Others go to 
the extent of saying that the activity of consciousness is the 
primciple of the word itself. For it has been said— 

“The word which has been taught in all the Sciences as 
the ultimate source assumes form through the appearance of 

The divisions of this word, cows and the like, made 
known through the words , 3 having their source in the word, 
are not beyond the unity of the ultimate even though they 
appear to be different. " 

K. S. 15 




They conquer death who know the word which has six 
doors, six bases, six forms of knowledge and six eternals. 4 

[1. Samjna, sasamjna, visamjna , antahsamjna, bahihsamjna, these 
five allied expressions are found in this verse and the Vrtti thereon. 
Visamjna is also found in the verse quoted in the Vrtti. Sasamjna and 
Visamjna mean sentient and insentient respectively. A piece of wood 
is given as an example of what is Visamjna. Antahsamjna and Bahih - 
samjna seem to stand for different states of beings: when conscious¬ 
ness is withdrawn from external objects it is antahsamjna, when it is 
turned outwords, it is bahihsamjna. 

2. Svaparasambodhah. Whatever has consciousness is capable of 

svasambodha or both svasambodha and parasambodha _whatever is 

sthavara , stationary, like a tree has only svasambodha. Men and other 
animals have both, according to Vrsabha: svasambodhdnugama eva 
sthavaresu, ubhaydnugamo jahgamesu manusyadisviti. l Parasambodlia > 
can mean both knowledge of others and making others know. 

3. Vdnnetvdh vannibandhandh. Things consisting of objects and 

words, are brought to the mind through words ( vacanan niyante _Vr.) 

So they are called vannetrdli. In other words, we see everything through 
the word. S 

4. The v/ord is said to have four sets of six things. What thev 
stand for is not clear. Vrsabha explains them on the basis of the six 
kinds of Pratibhd mentioned in Vak. II, 154. due to (1) svabhava * 
(2) carana; (3) abhyasa; (4) yoga; (5) adrsza; (6) visistopanatd. Thev 
are the means ( dvara ) of attaining the ultimate word, the six mean¬ 
ings which figure in them are its basis (adhisthana) , the six kinds f 
cognitions which result from them lead to it and the six relations with 
tne six meanings are eternal (avyaya) .] 

127. Just as in the wakeful state ( pravibhdge ) 
it is through the word that the agent acts on the ob- 

J ' ects ° f a11 acti ons, in the same way, in dream 
(avibnage), it is the word itself which becomes the 
object of all actions. 

. . 'P le a PP eai ’ances of the Word-Brahman (in the wakeful 
e consist of things to be accomplished and the means 

the divisions Wakeful state > the see <k of 

tne word, being of the nature of the word, 

I. 128. 


attain modifications like birth etc. through production, modi¬ 
fication and reaching and act as objects of action. In the 
undivided states like sleep, on the other hand, the same word, 
being deprived of (external) objects but being similar to them, 
becomes the object of the actions of production, modification 
and reaching. 1 It has been said— 

The Lord of All, the All-Embracing, the Enjoyer, after 
dividing Himself and after having created many different 
things, proceeds to sleep.” 

[1. See Vak. Ill, Sa 45.] 

128. Whether everything is of the nature of the 
Self or of the Supreme, as the word presents it, so it 
is understood; it is through the word that the object 
is established. 

.. ng !° T- a “ is a product of 

Ac Self. It exists within in every individual, but appears to 

be external. That something is Internal and something else 
is external is mere usage, based on rooted previous practice. 
But this is impossible considering that the source of both is 
one and incorporeal. According to others, all forms of know¬ 
ledge and all differentiation is a transformation of the inner 
consciousness and so on. Such are the views of those who 
hold that everything is a product of the Self. Others, how- 
ever, think that the Supreme Self . 

which emerges out of it as oil does from the sesame seed. 
Others still think that the process is like the production of 
sparks from fire of masses of cloud from subtle air, of streams 
of flowing water from the moon-stone, of sala trees and the like 
from the earth, of banyan trees with their downward growth 

l r °m + , an / an See ? and S ° ° n ' Such are the views of those who 
hold that everything is a product of the Supreme Self. The 

views of those who are for the Self and these who are for 
the Supreme have to be understood from the Philosophical 
commentaries. What is meant by ‘as the word presents it’ 
is this — the word which is within is presented as pleasure 



or pain in many ways in different beings. Even in such cases 
as the throwing and falling of stones, 2 it is so presented accor¬ 
ding to worldly usage and established practice. What is meant 
by ‘it is through the word that the object is established is — 
It is the word which creates the object and preserves it. 

[1. The views expressed in this verse and the Vrtti are not those 
of Grammarians. It is intended to show here that many views are 
just creations of words and do not correspond to reality. There are 
some who hold that everything is a creation of the individual Self 
while others hold that everything proceeds from the Supreme Self. 
They are respectively called svamdtrdvddinah. and pdramdtrdvddinah . 
Among those who hold that the universe is a product of the indivi¬ 
dual Self, there are those who do not specify whether, by individual 
Self, they mean the Intellect ( Buddhi ) or the individual consciousness. 
As Vr puts it:— caitanyam ekam karanam, purvatra tvanirupito ‘ ntali - 
sannivesl purusa iti visesah. Among those who declare the Supreme 
to be the source of everything, those who give the emergence of oil 
from sesame-seed as example seem to think of gradual creation and 
those who give the other four examples, of simultaneous creation. 

2. Lostaksepapatddisu. Vr$abha gives reference to the following 
passage from M. Bha. I, p. 123, 1. 11. 

“Acetanesvapi. Tadyathd. Losth ksipto bdhuvegavi gatva naiva 
tiryag gacchati , norddhvam arohati , prthivivikdrah prthivbneva gacchat - 

In the course of the discussion on P. 1.1.50, where we are told that 
of many possible substitutes, the nearest to the original must be chosen, 
somebody argues that the word * antaratama 3 , in P. 1.1.50 is unnecessary 
as the nearest would be adopted in any case, that being the practice 
in the world, not only among living things, but also among insentient 
things. To illustrate how even insentient things associate with what is 
nearest to them, the example of a stone thrown upwards is given. 
After going up to a certain height according to the strength of the 
thrower, it does not proceed horizontally, nor does it go further up¬ 
wards, but being a product of the earth, falls and joins the earth below. 
Here, however, the throwing upwards of a stone seems to be mentioned 
in another context. The context is the idea that everything is under¬ 
stood by us and affects us as words present it. As Vr. puts it: — 
Prdnidharmesvdkhydya b&hyesvapi sabdasya vyap&ram aha lostaksepa 
iti” It is, however, a pity that the text of the Paddhati here has not 
come down to us faithfully, so that the connection between the Bhasya 
context and the present context, as understood by Vrsabha does not 
become clear.] 

I. 129-130. 


129. In the case of a thing like ‘a circle of 
fire’ (alatacakra) where the circumstance is totally 
different, merely by the force of the word, its form 
clearly figures in the mind. 

It is the audible word which shows all meaning as resid¬ 
ing in its own form, the word. It seems to create that mean¬ 
ing as it were. It always exists in it as something to 
be conveyed by it. The word is not concerned as to 

whether the object in question has real existence out¬ 
side or not, nor whether there has been a mistake or 

not. Even in the case of a circle of fire, the pre¬ 

sence of a word which bears resemblance to the one which 
denotes an action consisting of spreading in all directions, 
creates a meaning for practical words like alatacakra (circle 
of fire). The form of the object so fictitiously created, be¬ 
comes conventionalised, even in the face of strong inf erence 
to the contrary. In the case of objects like a hare’s horn, 
definitely known not to exist, it is the audible word which 
creates or rather brings the idea thereof to the mind and 
binds it to its own form. So also in the case of objects which 
are accessible to the senses (pratyksari ). A meaning, whe¬ 
ther it has an external basis or not, is always present in the 
word, as the thing to be expressed by it. (Whether the object 
exists outside or not) in every individual, following the im¬ 
pressions of his previous experience and according to his 
knowledge, meanings of different kinds are understood from 

130. It has been said that the Self, which is with¬ 
in the speaker is the word, the great Bull with whom 
one desires union. 

Here (in the Science of Grammar) the word is of two 
kinds: it is eternal or it is a product. The product is that 
which is found in worldly usage and it bears the reflection of 
the Self 1 which is essentially the word. The eternal one is 
the source of all usage, 2 it has all sequence suppressed, it 


VakyApadiyam of bhartrhari 

resides within everybody, the source of all transformations, 
the substratum of all actions, the basis of pleasure and pain, 
unimpeded anywhere in regard to the production of effects 
(but) with its field of enjoyment restricted like a lamp cover¬ 
ed with a jar, the limitless source of all corporeal objects, 
manifesting itself as all forms of knowledge and as all diffe¬ 
rentiations, imitating the states of sleep and wakefulness 
through cessation of activity and resumption of it respectively, 
endowed like rain and forest-fire with the powers of produc¬ 
tion and destruction respectively, the Lord of All, endowed 
with all powers, the great word—Bull; those who know the 
process of union with the word break the ego-sense and are 
united with it, in complete absence of differentiation. It has 
been said 3 — 

He has four horns, three feet, two heads and seven 
hands. He, the great Bull, is tied in three places and roars. 
The great God has entered into human beings.” 

[1.^ Purusasya prabimbopagrahi. Vrsabha comments— ‘Yo’yam 
rathyapurusah so. vaktattvavikaratvat kdryasabdasvabhavah, vikara- 
V&m prakrtirupdnvayat = The man in the street is of the nature of the 
transitory word, because he is a transformation of the word-principle 
and the source persists in the products. 

2 . ^ Sarvavyavahdrayonih. According to Vrsabha, this means that 
1 f 1S ii f Source of the distinction between the end and the means and 
°' 1 the distinctions accepted in the different sastras. 

3- RV IV, 58.3.] 

131. Therefore the purification of the word is 
(the means to the) attainment of the Supreme Self. 
One who knows the essence of its activity attains the 
immortal Brahman. 

Once the essence of the word is purified by the form the 
correctness of which is established and particular merit is 
Manifested by the disappearance of obstacles in the shape of 
incorrect forms, well-being ( abhyudaya ) is certain. Through 
repeated practice of it and after attaining union (with the 
word-principle) through the correct word and after fully un- 

I. 132. 


derstanding the Intuition which derives from the word-prin¬ 
ciple and which is the same as being, which is the source of 
its modifications and possesses the powers of being the means 
of accomplishment and the thing to be accomplished, attain¬ 
ment of the Supreme Good (ksema) is certain. 

“After taking his stand on the word which lies beyond 
the activity of breath, after having taken rest in oneself by 
the union resulting in the suppression of sequence, 

1 Aftei having purified speech and after having rested it 
on the mind, after having broken its bonds and made it bond- 

“After having reached the inner light, he, with his knots 
cut, becomes united with the Supreme Light.” 

132. Nobody admits that there is any written 
tradition not associated with a particular author. When 
all such written traditions disappear, the three Vedas 
continue as the seed. 

In all systems, somebody is thought of as the author of 
the written tradition and so its human origin is accepted 
Vedic sentences, on the other hand, are like consciousness 
itself, not created by any person. When the authors of the 
written traditions will have perished, they will serve as the 
seed for the formation of other traditions.* 

[1. The word agama is used in a wide sense t„ , 

the Vrtti and the Paddhati refer to the MahabhSsva a - Y P aCeS T ’ bot 
Vrtti on verse 11, the word ‘dgamena’ occurs and it refers tTth m i!- 
bhasya. Vrsabha calls the twelve verses quoted fa the Vrtt 
1 at the end, the quotation suksmam etc. and v&maivedam J 'T® 
in the Vrtti on vereee 5 end „ respee.iveiy, 

quotations from the M. Bha and the Sahgraha, found fa the Vrtti h 
the name agama. (See the Vrtti on verses 23 and 26 and the pfddhal 
thereon.) ldl 

Here the word occurs both fa the verse and the Vrtti. It stands 
for ancient tradition, especially for written Tradition.] 



133. Even if the doctrines perish and there are no 
more authors to compose others, cultured people follow 
the right path mentioned in the srutis (Scripture) and 
the Smrtis (written tradition). 


In all discussions, it is admitted that, like the authors, the 
written traditions themselves can disappear. When they come 
to an end and before other authors arise and other written 
traditions are elaborated, there may be an interval during 
which cultured people do not violate the rites taught in the 
Scripture nor the regulations relating to what to eat and 
what not to eat embodied in the written traditions. 

134. If knowledge were spontaneous, there would 
be no purpose in the composition of the sastras. If 
spiritual merit is the cause of knowledge, the Veda is 
the cause of the former. 

If it is accepted that a particular person can acquire 
knowledge without instruction, then written traditions prohi- 
itmg what is harmful and enjoining what is beneficial would 
ecome useless. If only a particular individual, due to special 
merit, acquires knowledge without instruction, while some 
°t ers have to be instructed by a treatise, then the merit 
w ich has brought about that particular individual must have 
a e mite basis. All other bases disappear. Therefore, writers 
w o follow the right path which is based on Scripture, com¬ 
pose ifferent doctrines and attain different modes of reali¬ 
sation. ccilL 

TT , 135 ' Reas °ning which does not go against the 

‘ aS J md the sastra is authority for those who cannot 

P " . e mea ning of a sentence does not become clear 
a '°m its mere form. 

mss^rtT mnS * makeS divisions within the words and mean 

the benifit l* ,he Soripture ' A ” d that *> 

efit of those who see superficially. That is why sue! 

I. 136. 


reasoning has been adopted by the ancients in the investigat¬ 
ing sciences (nyayavidycLsu). Only that much of reasoning 
is to be adopted as will not go against one’s Scripture. In all 
the sacred works, there is much that is beyond reasoning and 
has to be accepted through faith. What purpose does he seek 
to serve who does not follow reasoning, but accepts Scripture 
alone as authority? Correct understanding of the sentences 
of the Scripture would be his purpose. The power of the 
sentences of Scripture to convey their meaning varies, even 
when their form is the same, due to the presence of some 
other factor. He who understands the meaning from the mere 
form, without relying on context, connection etc. ends in con¬ 
fusion as to what is meant and what is not meant. 

136. Sometimes what the words say is not meant, 
sometimes it is included in something wider, sometimes 
it is specified by other indications, many such conclu¬ 
sions are arrived at by reasoning. 

In the sutras 

“What the agent wishes most to reach is the object.” 
(P. 1.4.49) 


“The suffix has the meaning of ‘child of’ so and so.” 
(P. 4.1.92.) 

and in the sentence 

“He cleans the vessel” (Cf. Tai. Sam. and Jai. 
Mi. Su. 3.1.13-14.) 

and in the sutra 

“The person for whom the object is meant is the rece- 
pient.” (P. 1.4.32.) 
and in 

‘Those who live on a woman have attained the charac¬ 
teristic of a dead person’ 

JC. S. 16 


gender, number and tense are not meant; sometimes 
they are meant; such specification in regard to definitions 
depend upon reasoning. Similarly, in the sentence 

“He should release his speech after seeing the star. 
(Ka. Sam. 23.5.) 

the seeing of the star is meant to convey a particular time 
(when the stars are ordinarily visible). Thus, it, the releas¬ 
ing (of speech), is done when the main thing (the time) is 
otherwise ascertained or as (seeing the star) stands for some¬ 
thing else, the particular time is ascertained when the stars 
are visible. In the sentence let the curds be protected from 
crows,’ as it is meant to ward off all damage, even if there are 
no crows, the curds are protected from dogs etc. As the sen¬ 
tence let the vessels be cleaned’ is meant to denote the com¬ 
pletion of all the preliminaries to the act of eating, even if 
there are no vessels, the other preliminaries are completed. 
Clarifications can be done by other indications also ( linga ). 
In the sentence “he puts the sugar mixed with fat,” all kinds 
of fat come to the mind, but through the indication contained 
in the sentence— 

“Clarified butter is indeed lustre itself” 
a particular fat is understood. Such conclusions are reach¬ 
ed by reasoning and definitions are specified. 

137. Reasoning based on human intelligence (as 
distinct from written tradition) is also ultimately the 
power of words. Argument not based on woids among 
those who have no written tradition, is really without 
any basis. 

It is the word which is the instructor. Speakers follow 
the power of words and act when urged by a desire to speak 
based on the availability of the right word. When the hearer 
follows the fixed power of words to convey a meaning, 
through reasoning based on meaning, context, indication, 
interconnection etc., people look upon the capacity of the 
words as the reasoning of the hearer. 1 As for the reasoning 

L 138. 


which does not proceed from the power of words, but follows 
the similarities and dissimilarities of objects, it is destructive 
of all tradition, without any basis and it is called ‘dry rea¬ 
soning/ As in the following— 

“If the drinking of wine contained in a big circle of rose- 
coloured jars cannot take one to heaven, what can the little 
that is drunk in a sacrifice do?” 2 

Or in the following— 

“To say that one should express oneself in words means 
that one should do so in the Dramilaka language.” 

[1. The relevant considerations for determining the meaning of 
words are given in Vak. a. II, p. 214 ff. 

2. M. Bha, I, p. 3, 1. 3.] 

138. Just as colour and other qualities are seen 
to have powers in regard to particular effects, so are 
words seen to have such powers to remove poison etc. 

For those who have understood from the sacred books 
that divine happiness results from the knowledge and use 
of correct words, accompanied by a knowledge of the science 
of Grammar, this verse provides an argument strengthening 
the tradition recorded in the sacred books. It is seen in this 
world that colour, taste, smell and touch, whether singly or in 
combination, have powers to produce particular visible or 
invisible results. 1 Power to produce visible result is seen in 
poisonous plants, magnet and trees. Power to produce invi¬ 
sible result is seen in wine and in the waters of holy places. 
In the case of some words also, visible result such as the 
curing of snake poison is seen. Similarly, it is accepted that 
an invisible result is obtained by the repetition of mantras. 

[1. Vrsabha gives the following examples. Blue and bright are both 
colours, but one is soothing to the eyes and the other is irritating; sweet 
and bitter are both tastes, but one causes phlegm and the other bile. 

These are visible results. Only a white goat can be sacrificed to 
Vayu, otherwise the invisible result expected will not be produced. The 



touch of wine leads to a bad invisible result while that of the waters 
of holy places leads to a good one. That the mantras of the Atharvaveda 
have the effect of curing snake-poison is an example of visible effect 
from the word.] 

139. Let it be understood that just as they (the 
words) have such powers, similarly they also have the 
power to produce merit. Therefoi'e, good people who 
desire well-being should use only correct words. 

It is learnt from the sacred books that there are powers 
leading to visible results associated with every word. As 
there is such a tradition, let it be understood that correct 
words are the means of acquiring merit and, therefore, those 
who wish to attain divine happiness must follow the trust¬ 
worthy tradition uninteruptedly practised by the cultured 
and use only correct words in worldly transactions. 

140. Everybody understands things having invisi¬ 
ble effects from the sacred texts. In regard to every¬ 
thing mentioned in it, it would be possible to postulate 
the opposite. 

Some argue as follows:— Just as words have the capa¬ 
city to cure poison, in the same way, why not understand that 
they have the capacity to cause demerit. From this very 
illustration, let one conclude the existence of an opposite 
power. Against this objection, it is maintained that, in regard 
to all conclusions relating to visible or invisible results of 
sacred texts, it would be possible to postulate an opposite 
effect. Therefore, if after taking some sacred text as autho¬ 
rity, its meaning is settled, then any reasoning that is put 
forward would strengthen the understanding of it. 

141. Knowledge of the correctness of words is 
the subject of this tradition called Grammar. It is here 
that the uninterrupted tradition of cultured people is 

I. 142. 


Just as traditions relating to what can be eaten and what 
cannot be eaten, which woman one can marry and which 
woman one cannot marry, what can be said and what cannot be 
said are well established and cultured people do not go against 
the code of conduct based on them, in the same way, this tra¬ 
dition called Grammar relates to what particular words can 
be used and what not. What is remembered from generation 
to generation, in an uninterrupted manner is again and again 
embodied in woids. A tradition which has no written basis 
but the observance of which is well-known is preserved by 
the continuity of the practice of the cultured. 

142. This Science of Grammar is the supreme 
and wonderful source of the knowledge of the three¬ 
fold word, comprising many paths, of the Vailchari (the 
Elaborated), the Madhyamd (the Middle One) and the 
Pasyantz (the Seeing One). 

That .s caUed the Elaborated (Vaikhan) the form of 
„h,ch is cognised by others inasmuch as it comes within the 
range of the sense of hearing, and is well defined. It is mixed 
up or uttered indistinct phonemes; it has well-established 
eorrectness or ,s devoid of correctness. It is what comes out 
of the axle of a cart a drum, a flute and a lute («). th u s 
it has minute varieties,. The Middle One (UaihyaL), on 
the othei hand, is what exists within, it looks as if it has seau- 
ence and mind is its only substratum. According to some it 
is accompanied by the subtle functioning of breath (prfl) 
and even though sequence is suppressed in it, yet it Tas 
distinct functioning of breath in it 2 The Seeing Qne (Pos ,_ 

yanti) is'that in which sequence is merged and though it is 
One, the power to produce sequence has entered into" it. It 
is restless (caUcald) and also still in concentration^ hidden 
and pure; 4 the forms of the objects of knowledge have entered 
into it or merged into it or it has no form at all; it has the 
appearance of limited objects or of connected objects or the 
appearance of all objects has come to an end in it; thus it has 
infinite variety. According to some, in all the states of the 

126. vakyApadiyaw of bhartrhabi 

word which come within the range of usage, the distinction 
between what is correct and what is not is well-established 
and brings about the perfection of the individual. But the 
supreme form of the Seeing One is devoid of a 11 correct forms, 
it is not mixed up and it is beyond worldly usage. According 
to the tradition of some, it is the form of this (state of the 
Word) which can be attained either by a knowledge of Gram¬ 
mar or through Union preceded by the word, obtained by the 
knowledge of the correct form of the words. 5 On this point, 
illustrations are found in the Itihfisci. G 

'of them, the word—Gow, having a ladiant smile 
and an excellent nature, in her divine and non-divine form, 
yields milk as the Cow does.” 

“See the difference between the two, both subtle and 
throbbing; the other one is present in the interval between 
prana and apana.” 1 

“Another one, not being prompted at all, exists without 
the breath. From that is breath born, which, in its turn, 
strengthens the word.” 

“Strengthened by the breath, it becomes the basis of 
usage; by reaching the breath of everybody, the word does 
not speak in any case.” 

“That which has always resonance, that which has 
acquired resonance, and that which has no lesonance at all, 
exist. To the two which have resonance, that which has none 
at all is superior.” 8 

Again, it has been said— 

“The Elaborated (Vaikhan) assumes the form of phone¬ 
mes when the air strikes against the places of articulation 
and is based on the functioning of the breath of the speakers.” 

“The Middle One proceeds by going beyond the function¬ 
ing of breath, has sequence and has the mind alone as its 

“The Seeing One is indivisible and is entirely without 
sequence. It is the Inner Light, the subtle word, imperish¬ 

I. 142. 127 

“Though always attacked by adventitious impurities, it, 
like the last digit of the moon, is never overwhelmed.” 

“When its form is perceived, all obligations cease. 9 Of 
man having sixteen parts, that is the immortal part.” 

“The Elaborated One ( Vaikhari ) coloured by adventi¬ 
tious impurities, is not really affected by them, any more 
than the pure Being is affected by qualities.” 

This three-fold word, has, like the appearances of the 
knots of Consciousness, indefinable dimensions, and a fourth 
of it exists in man. Even of this, only a little comes within 
the range of usage, the rest is beyond the usage of ordinary 
man. It has been said— 

“There are four states in regard to the word. The wise 
among the Brahmanas know them. Three of them are kept 
in a cave and do not move. Men speak the fourth part of the 
word.” 10 

The pure form of that word is embodied in this descrip¬ 
tive Science called Grammar, consisting of general and 
special rules. The powers of those of lower visions, generally 
admit of degree, are liable to meet obstruction and to com¬ 
mit mistakes. Hence this method of acquiring the words, free 
from error and consisting of definitions and elaborations and 
containing many paths, has been developed. 

[1. Slista vyaktavarnasamuccdrana prasiddhasadhubhdva bhrasta - 
sarnskard ca ... apariviaiiabhedd. 

It is stated that Vaikhari has infinite varieties. To an objection that 
what comes out of the axle of a cart is just noise and not vdk 9 Vrsabha 
replies that the inclusion of it is based on a sruti which he proceeds to 

By explaining slista ns avyaktdksavd, he makes it the opposite of 
vyaktavarnasamuccaraiia. It is not clear whether the noise coming out of 
the axle of a cart or a drum is meant to be an example of slista. If 
even such noises can be examples, it is only natural that incorrect forms 
of words also should be considered as varieties of it. They are, after 
all, Vdk. 



2. Kramasamlvarabhdve 3 pi vyaktapranaparigrahaiva. Vrsabha gives 
as a proof of the presence of subtle prana in the Mahyama stage that 
sequence can be observed when one silently recites something to one¬ 
self. As he puts it— Svayamapyasyopdmsutaram pathatah tat spastam . 

3. Caldcala pratilabhasamddhdna ca. Vrsabha explains these two 
expressions in terms of Yogic teminology. He says: calacalaiti rupadisv, 
visayesv arvagdarsananam viksiptotpadyate buddhir, vageva hi sd. Pra- 
tilabdhasamddhand ca iti aviksiptd Yogindm sabdapurvayogena samdhi- 

4. Vrsabha explains visuddha in two ways: (1) free from all diffe¬ 
rentiation, as Yogis see it, (2) free from all corrupt forms, as Gram¬ 
marians know it. 

5. According to some, the supreme form of the Pasyanti can be 
attained in either of two ways: either through a knowledge of Gram¬ 
mar or through union with the word, induced by a knowledge of the 
correct form of the word. Vrsabha also understands it in this way. 
The use of the expression ‘ekesam’ suggests that this is not the ortho¬ 
dox view. In fact, elsewhere sabdapurvayoga alone is mentioned as 
the means of attaining it. See the Vrtti on Vak. I. 14, 131. 

6. The following verses are found in a very different form in the 
Mahabharata, Asvamedhikaparvan, 22. 

7. Pranapanantare. Vrsabha explains f antara 9 as absence. So the 
absence of prana and apana would mean something beyond the range 
of prana, ie, buddhi, the seat of the Madhyamd. 

8. In these verses which are supposed, to be quotations from Iti - 
hasa, ie, the Mahabharata, the names Vaikhari , Madhyamd and Pasyan¬ 
ti are not used (unlike the later quotations where they do occur), but 
the descriptions are sought to be interpreted by Vrsabha as relating to 
them. Thus pranapanantare tisthati is equated with Madhyamd,; dpur- 
yamaneva vina prdnena tisthati is also understood as referring to Madh¬ 
yamd, vyavahdranibandhana means vaikhari. Ghosini and jatinirghosd 
are interpreted as standing for Vaikhari and aghosd for Pasyanti and 
Madhyamd together. 

9. Adhikdro nivartate = all obligations cease. Adhikara is explain¬ 
ed by Vrsabha as niyogo bandhdkhyah. The bond of obligation which 
is the cause of samsdra ceases to be and one attains Moksa. 

10. R. V. I, 164, 45.] 

143. As the powers of words are seen by those 
who know the true nature of things, the Science of 

I. 144. 


Grammar is composed on the basis of analysis or with¬ 
out analysis. 

What is called analysis is the postulation of divisions like 
stem and suffix for teaching others. For example, the state¬ 
ment that the suffix tavyat etc. are added to the root. So it 
has been said— 

“Whatever he can teach in a general manner, he does 
accordingly.” 1 

There is absence of analysis where a word is given 
as it is. For example, the words dddharti, darddharti, 
etc. 2 and the words dasvdn, sahvdn? In some Grammars, 
there is a great deal of absence of analysis and they teach 
many words directly (pratyaksapaksena ). Others, on the 
other hand, resort to analysis and by adopting the indirect 
method of inference, teach whole groups of words. This Gram¬ 
matical tradition is established in different ways, according 
to the times and keeping in view the mental capacities of 
people. There are cultured people and that they are so can be 
inferred only from their use of correct words. They, whose 
inner vision is unobstructed in regard to all things to be 
known, see, without error, the power of words, differing at 
different periods of time and consisting in their being the 
means of merit or demerit, according to the times. 

[1. M- Bha on P. 1.1.46. 

2. P. 7.4.65. 

3. P. 6.1.2.] 

144. Scripture ( Sruti ) has been declared to be 
beginningless, continuous and without an author. 
Written tradition ( Smrti ) is composed by cultured 
Ancients and has continuity. 

There are those who accept the authority of Scripture 
only and consider only the sacred books as trustworthy in 
regard to matters having invisible fruit and the views of men 
as doubtful and, threfore, not authority. According to them, 
even though there is no difference between Scripture and 
K.§. 17 



written Tradition as far as continuity of what is taught is con¬ 
cerned, yet Scripture has no deviations in the rules relating 
to accent, phoneme, sequence, place and time of study nor 
has it been established differently by any other person and 
it has always been established in all the countries in divisions 
according to branches. Written Tradition, on the other hand, 
has continuity of meaning, but is composed by the cultured 
differently at different times in prose, verses, sentences etc. 

Some teachers think as follows: No act has, in itself, a 
visible or invisible fruit. It is only by acting according to 
Scripture that merit is manifested and by going against Scrip* 
ture that one is tainted with sin. Scripture itself ordains that 
an act like the killing of a Brahmana which is a sin in some 
contexts becomes the cause of divine happiness if done in 
another context. 1 

Others, on the other hand, think that Scripture only 
makes known the specific power of objects. What would be 
the interest of Scripture to be a playful cause (of merit and 
demerit) and favour or harass men? It is seen to be better to 
assume that it (the causing of merit or demerit) is the nature 
of substances rather than that it is the nature of Scripture. 
In such written traditions as that of the Science of treatment, 
it is things like poison or herbs which have the capacity to 
fulfil a purpose and not the traditional texts. Therefore, it is 
nature of substances which is followed by Scripture, as it is 
the nature of Scripture in the other view. Just as, in the 
world, causes of benefit or harm are clearly distinguished, 
in the same way, there is uninterrupted establishment of the 
means of knowing them. 2 

[1. Vsrabha points out that in Purusamedha and Sautramani, braJi- 
manavadha leads to heaven. As he puts it— Purusamedhe sautrdman - 
yam ca svargangatvam. 

2. What is meant is that the power to produce their effects, good 
or bad, belongs naturally to substances, but, by uninterrupted tradi¬ 
tion, it is Scripture that makes it known. Similarly, the power of 
correct words to produce merit and that of incoirect words to produce 
demerit are inherent in them. Grammar only makes them known]. 

i. 145. 


145. In those who evolved out of the un¬ 
differentiated, there is knowledge of the Sruti (reveal¬ 
ed Scripture) as in a dream. The Written Tradition, on 
the other hand, is composed by the sages, after under¬ 
standing the nature of things and following the indi¬ 
cation (found in the Vedas). 

It has been stated in the immediately preceding verse 
what the dliarma is that is taught in the Scripture and in the 
written Tradition according to those who believe that the 
present diflerentiated condition of the world is eternal, that 
distinctions such as Yuga, Manvantara, etc. do not exist nor 
such an extraodinary division called the day of Brah¬ 
ma. According to those who hold that the ultimate cause 
works in the manner of sleep and wakefulness and the diffe¬ 
rentiated individuals, some sages manifest themselves as iden¬ 
tical with Intuition; they see it, the great Self in the form of 
Being, the source of Nescience and endowed with all know¬ 
ledge and they become one with it. Some sages manifest them¬ 
selves together with (the means of) knowledge. They iden¬ 
tify themselves with their Self in the form of mind-knot, free 
from the elements ether etc., either severally or collectively, 
ie., devoid of any sense of T in regard to them. All the activi¬ 
ty of those sages is the product of Nescience and, therefore, 
adventitious and secondary. Their being essentially know¬ 
ledge is eternal, non-adventitious and primary. They see the 
whole Scripture, endowed with all power of differentiation 
and all power of unity, as one hears sound in a dream, inaudi¬ 
ble to the ear. 

Some other sages, after perceiving the nature of objects, 
conducive either to the welfare or to the harm of man and 
after seeing in the Scripture indications thereof, compose the 
tradition, (the observance of which) leads to visible and 
invisible results. At first, they hand down the Scripture in 
an undivided manner, without any deviation in the words, as 
they saw it and later, they hand it down, divided into bran¬ 
ches. Such is the tradition. 


146. The impurities which belong to the body, 
the speech and the mind are removed by the sciences 
of Medicine, Grammar and Philosophy. 

Just as, after seeing forces leading to illness in the body 
and the capacity of precious stones, herbs etc. to remove it, 
the Science of Treatment has been initiated; just as, after 
seeing that passions etc. cause disturbance of the mind, 
systems of philosophy have been composed which are the 
means of acquiring knowledge which leads to their cure, in 
the same way, the Science of Grammar has been initiated in 
order to make known the features of correctness in words and 
to enable us to discard the corrupt forms which are obstacles. 

Of what nature are the corrupt forms? 

147. When one wants to utter the word gauh and 
actually utters one which is devoid of correctness to 
convey that particular meaning, it is called a corrupt 
form ( apabhramsa ). 

The author of the Sahgraha has said: 

“The correct word is the original of the corrupt form.” 

There is no corrupt form which is independent, with¬ 
out an original. Of every corrupt form, a correct form 
is the original. But some corrupt forms acquire inde¬ 
pendence by being constantly used and becoming con¬ 
ventional. Instead of saying gauTi , one uses, either through 
lack of capacity or through mistake and other causes, the 
corrupt forms gcivi etc. having the correct form as the origi¬ 
nal. Such forms lose their correctness when used to denote 
an object having dew-lap, etc. When used in another meaning 
(which is its own) it is looked upon as correct. Their incor¬ 
rectness is not linked with their form only. 

148. Words like ‘asva and ‘gom’ are correct in 
regard to other meanings. Everywhere correctness is 
determined by the particular meaning which the word 
is to convey. 

1 146 . 


The two words genii and asva are correct, the former in 
the sense of a receptacle and the latter in the sense of one 
who owns nothing. These two words which have acquired 
their correct form in regard to some other meanings are con¬ 
sidered to be correct if, due to some special reason,! they are 
used to denote an object having dew-lap etc. and one which 
neighs etc. respectively. When a cow carries much milk and 
is, theiefore, like a receptable, one says: this cow is a recep¬ 
tacle because she is like a receptacle. In the same way, a 
horse {asva) is called ‘asva’ when it is a priceless one. 1 
Therefore, till some feature is isolated on the basis of which 
correctness or incorrectness can be established, no object is 

[ 1 . Tatlid* sasnddimati hresitddilinge ca .... tathdvidyamdnam 
svam asya so’yamasva iti. What is stated here that, in certain circum¬ 
stances, the words ‘goni’ and ‘asva’, applied to a cow and to a horse 
respectively, would be correct. Figurative use of words is meant here. 
This passage of the Vrtti should be compared with the commentary 
of Bhartrhari on the M. Bha (Paspasahnika), p. 13 (the edition now 
being printed in Benares) or p. 11 (of the edition of K. V. Abhvankar 
and V. P. Limaye) A.B.O.R.I., Vol. XLIII: 

Sa eva sabdo’rthavisese Kasminscidasddhurityetaddkhyayate. 
Yatha gonisabdah sasnddimatyasadhustatha asvasabdah kesarddimati 
sddhur na nihsva iti. asva iti nirdhane sadhuli, naikasapliddilaksane. 
Yadi tu gonisabdo’pi nimitt&ntarat sasnddimati prayujyeta goniva goniti 
sddhur eva syat. Asve va asvasabdairi dhan&bhdvadvdrakam prayunjxta 
s a sddhur eva.] 

149. They (the incorrect words) are the cause of 
the correct words (figuring in the mind) by inference. 
By identifying themselves, as it were, with them, they 
convey the meaning of the correct words. 

When corrupt forms are used to denote objects which 
are the sphere of the correct forms, they convey the mean¬ 
ing through the correct words, just as winking etc. are, 
through familiarity, well known to appear as having taken 
0 n the form (of the words which express the same idea). 



The purpose of another verse 1 on the subject is to say that 
corrupt forms do not convey the meaning directly. 

[1. That corrupt forms do convey a meaning was already stated 
in Vak. I. 147. Visistarthanivesinam.] 

Why is it that gon\ etc. are not considered to be syno¬ 
nyms of gauli? In regard to such matters which are embodied 
in the written Tradition, no other cause other than being 
well-known in the practice of the cultured can be adduced. 
If gavl etc. were synonyms, they would also have been in¬ 
cluded in their rules and actually used. That word is correct 
which is used, like perception, to convey the thing which 
cause its use. The correct words convey that meaning which 
directly prompt their use. Therefore, it is being said— 

150. Since they are not explained like correct syno¬ 
nyms by cultured people in the written Tradition 
(smrtisdstra ), therefore, they are‘not directly expres¬ 

The meaning of the verse has already been explained. 

151. Just as the child who is learning the words 
amba, amba, pronounces them indistinctly and those 
who know his speech understand their distinct forms, 

When the baby makes the necessary effort, but due to 
deficiency in the vocal organs, utters indistinct sounds when 
it wants to use the correct word, those who hear it understand 
the distinct word which is the original thereof. And they 
look upon the correct word as connected with the meaning 
and not the corrupt form used by the baby. 

152. In the same way, when an incorrect word is 
uttered in place of the correct one, the understanding 
of the meaning is preceded by the correct word coming 
to the mind. 

I. 153-154. 


When words are mixed up, corrupt forms are used 
instead of correct ones. But the cultured people who know 
Grammar understand the correct ones. They also look upon 
the meaning as having been expressed by the correct words. 
The corrupt form is the cause of the inference of the correct 
one as smoke is of fire. 

153. In the case, however, of those ignorant 
speakers among whom incorrect words have become 
established through habit, the correct words do not 
(first come to the mind and then) convey the meaning. 

Corrupt forms are used habitually by women, sudras 
and candalas and by speakers who make mistakes and 
thus they enter into convention and usage is more based 
on them (than on the others). When a doubt arises 
after the use of the correct form, now it is resolved with 
the help of its corrupt form. They look upon the corrupt 
form as being on the side of perception and place the correct 
form on the side of inference. 

154. This Divine Speech has been mixed up (with 
the corrupt forms) by incapable speakers. Those who 
consider words to be transitory are mistaken in this 

It is heard that, in the old days, the word of those whose 
body itself was light was free from all corrupt forms as it 
was free from falsehood. But it became contaminated and 
through the persistence of the tendency towards contamina¬ 
tion by the repetition of the first corruptions, it became, 
for those speakers, the original and the convention. Up¬ 
holders of non-eternality, on the other hand, do not believe 
that correct words lead to merit and look upon the rules 
relating to correctness as similar to the rules among wrestlers 
and consider the whole collection of correct words as 
prakrta, i.e. derived from prakrti (the original, i.e. the cor¬ 
rupt forms). The form which is settled by some narrow- 



minded persons by regulating accent, correctness and so on 
is a modification and established later. 

155. Both (the correct and the corrupt forms) 
have been handed down to us uninterruptedly. Still, 
when, intending to use one (the correct form) the 
speaker uses the other (the corrupt form) it is not the 
latter which must be deemed to be expressive of the 

Even according to those who do not believe in old times 
nor in an unmixed Divine Word which existed sometime 
ago, this distinction between correct and corrupt words, as 
the distinction between which woman one can marry and 
which not, has always been transmitted, without a break, 
by the cultured. The word, which, like the talk of children, 
is actually used to convey a meaning when some other word 
was meant to be used and has entered into usage and the 
one (which is similarly used) but which has not entered into 
usage, neither is expressive of the meaning. In such cases, 
either the meaning is understood preceded by the remem¬ 
brance of the correct word or, as with the ignorant, something 
is understood from those words as from winking and the 

Thus has ended the chapter relating to Brahman, en¬ 
titled “Collection of Traditions” in the ‘Treatise on the 
Sentence and the Word’ composed by Harivrsabha, the Great 


In the previous chapter, the nature and purpose of the 
Word were briefly stated. It was stated, in a general way, 
that the Word is expressive of the meaning. There is difference 
of opinion as to whether it is the individual Word or the 
Sentence which is expressive. The purpose of this chapter is to 
expound in detail the nature of the expressive sentence. 

1 - 2 . In regard to the Sentence* opinion is 
widely divided among thinkers inasmuch as they 
declare it to be (i) the Verb, ( 2 ) the collection of 
words, ( 3 ) the Universal inhering in the collection 
of words, ( 4 ) the One indivisible Word, ( 5 ) the 
Sequence (of the words), ( 6 ) the Unification in the 
mind, ( 7 ) the first Word, ( 8 ) each Word requiring 
the others. 

[These are the eight alternative ways of looking at the 
Sentence current among thinkers :—The Universal inhering 
in the collection of words, the One indivisible Word, the Uni¬ 
fication in the mind, these three definitions of the Sentence 
come within the view that the Sentence is indivisible ( akhaiida - 
pakfa). The Verb, the Sequence of the words, the collection 
of words, the first Word and each Word requiring the others— 
these five definitions come under the view that the Sentence 
has divisions (sakhcindapak$a) . That the sentence is the collec¬ 
tion of words or that it is their Sequence, these two views are 
held by those who accept abhihitdnvayavdda = the view that the 
sentence-meaning is the inter-connection of the meanings con¬ 
veyed by the individual words. That it is the Verb or the 
first Word or each Word requiring the others, these three views 
are held by those who follow anvitabhidhdna= the view that 
the word conveys a meaning already potentially connected 
with those of the others. These two theories result from inter¬ 
preting the statement padaprakrtih samhita (Rk-pratiSakhya 11,1.) 
as a tatpurusa or as a bahuvrihi . That the Verb is the Sentence 
is explained much later, in verse 326. The eight verses from 


41 onwards explain that the collection of individual words 
constitutes the Sentence. Verses 7-14 and 19-27 explain that 
the indivisible Sentence is the sphota. Sphota is of two kinds: 
External and Internal. The former is either the Universal 
or the Particular. Five verses from 49 onwards expound the 
view that the Sequence of the words is the Sentence. Verses 
47 and 48 elucidate the view that the first Word or each Word 
requiring the others is the Sentence. The Mimamsaka defi¬ 
nition of the Sentence would come under the sanghata (collec¬ 
tion) view. Definition of the sentence naturally leads to state¬ 
ments on the nature of the sentence-meaning. If the verb 
constitutes the Sentence, then the sentence-meaning is in the nat¬ 
ure of Action. If the collection of words or their Sequence con¬ 
stitutes the Sentence, then its meaning is the interconnection 
of the meanings of the individual words. If the first Word or 
each Word requiring the others is the Sentence, then the con¬ 
nected meanings make up the sentence-meaning. According 
to some, the sentence-meaning is in the nature of purpose 
( prayojanam ). Vidhi , niyoga and bhavana , mentioned by others, 
are not different from Action ( Kriya ) and so they are not given 
separately here. The Buddhist view of the Sentence is very 
near to Unification in the mind {buddhyanusarrihrtih) and their 
view of the Sentence-meaning is very near to that accepted in 
this work, namely, that it is in the nature of Intuition ( prati - 
bha). The Nyaya view of the Sentence and Sentence-meaning 
would amount to samghdta and samsarga respectively.] 

The author now discusses whether the definition given by 
the author of the Varttikas and by the Mimdmsakas would agree 
with the above definitions or whether they have a different 
scope altogether. 

3. With the definition of a sentence given in 
this sTistra for regulating loss of acute accent etc. 
the [Mimamsaka) definition does not fully tally. 

[Katyayana gives the following two definitions of a sentence: 
dkhydtam sdvyayakdrakavi$e$anam vakyam and ekatiri vakyam (va 
9, 10. M.Bha.I, p. 367, 1. 10,16). The Mimamsaka definition 
is contained in Mi.Su.Il.\A2 :—arthaikatvad ekamvakyamsakan - 
k$am ced vibhage sydt. It is referred to in verse 40. In ay am 
dando haranena, there is syntactical connection (samarthya ). 

II. 4-5 


Therefore there should be loss of acute accent ( nighata ) in 
hara by P. 8.2.28 but that is not desired because, according to 
the definition of Katyayana given above, there are here two 
sentences and nighata takes place only if both the words, the 
nimitta and the nimitti are in the same sentence. In nadyds 
tiffhati kiile , Sdlindm ta odanam ddsydmi , there is no syntactical 
connection between the first two words and yet nighata in the 
first case and substitution of te for tava in the second case take 
place, because the two words are in the same sentence. To 
decide whether the two words are in the same sentence, one 
should know what a sentence is and Katyayana tells us what 
it is in the vdrttika quoted above. If we follow the Mi. defi¬ 
nition of a sentence, there would be nighata in hara in the sentence 
quoted above and that is not desired. So here the two defini¬ 
tions do not tally. In nadyds ti$thati kule> they do.] 

The Mimdmsaka definition is now referred to. 

4 . A sentence is a collection of words which 
in isolation require one another for particulari- 
sation, which, as a whole, do not require an 
outside word, in which the verb is the predominant 
word, which has qualifying words and which serve 
one purpose. 

[In this definition, the verb is said to be the most important 
word, but not the only word. That is why the sentence is said 
to be gunavat . The verb is also essential. Otherwise, nilo 
ghatah would also be a sentence. Unless the words require 
one another, they would not form a sentence. All this is an 
elaboration of the idea contained in Mi.Su.ll. 1.42.] 

5 . The word in the vocative case in the sentence 
vrajani devadattal (=let me go, Devadatta!) is a 
qualifying word. That being so, loss of acute accent 

[In Katyayana’s definition of the Sentence there is the word 
vi$e$ana which includes kriyavi§e$ana also. So in vrajani deva¬ 
datta !, the word devadatta ! which is in the vocative case is 
a kind of vise$ana of the verb vrajani. So it is in the same sen¬ 
tence and the suppression of the accent takes place according 
to the rules.] 



6 . Just as many words ending in the suffix ktva 
can qualify (that is, can be subordinate) to a verb, 
in the same way, they declare, a verb can qualify 
another verb. 

[In Katyayana’s definition of the sentence, the word akhya - 
tam is in the singular number. That means that there can be 
only one verb in a sentence. There can be more than one 
verb if one of them is the main one and the others subordinate 
to it. Just as many words ending in the suffix k ! va can be sub¬ 
ordinate to one verb and be in the same sene mce, in the same 
way, more than one verb can be subordinate to the main verb 
which follows them in the same sentence. That being so, 
there would be nighata in the verb vrajati in the sentence: 
purvam snati pacatitato vrajati tatah, because thatis the main verb. 
The siitra : tinatinah (P. 8.2.2S.) also implies that there can 
be more than one verb in the same sentence. More than one 
verb can result in more than one sentence if they are indepen¬ 
dent of one another. Thus Katyayana’s definition is not too 

7 . Just as the one entity, comprising the cog- 
nition of all objects is differentiated on the basis 
of distinction in what is cognised, so is the cognition 
of the sentence-meaning. 

[Bhartrhari really wants to set forth the following view: 
The sentence is the sfihota, either external or internal. It is 
external when it is clearly uttered. Till then, it is internal. 
In any case, it is indivisible. It has two aspects : the sound 
aspect and the meaning aspect which are identified with one 
another. It is essentially in the nature of knowledge or con¬ 
sciousness because it illuminates an object. Because of arti¬ 
culation, it assumes the form of sound. Though indivisible, 
it appears to have divisions just as our complex cognition, 
though one, appears to have inner differentiation because of 
the objects in it. The picture is one but we seem to see differ¬ 
ent colours within it. That is what happens with the sentence 
and the sentence-meaning. Both are indivisible like the 
flavour of a cold drink, or the juice in a pea-hen’s egg, or the 
form of a picture, the narasimha, the gavaya and our perception 

II. 8-12 


of a picture. The indivisible sentence is the sphota and the 
indivisible sentence-meaning is pratibha . But both appear to 
have divisions.] 

8 . Just as the One picture is explained through 
its different colours like blue which belong to its 

9 . in the same way, the One sentence, entirely 
self-sufficient and complete, is explained through 
individual words which require one another. 

[A colour-pattern is one but it is explained in terms of the 
different colours which figure in it. In the same way, the 
sentence is one and self-sufficient but it is explained in terms 
of individual words which appear to figure in other sentences 
and which require one another.] 

10 . Just as a word is analysed into stem, suffix 
etc. in the same way, the analysis of a sentence 
into individual words is also described. 

[In the varttika : na va padasyarthe prayogdt (Va.19 on P. 
1.2.64.) it is stated that it is the word as a whole which conveys 
the meaning and that its division into stem, suffix etc. is arti¬ 
ficial. In the same way, the division of a sentence into indivi¬ 
dual words is artificial. The words have no real existence.] 

11 . Just as one sees in phonemes parts resem¬ 
bling other phonemes, in the same way, one sees 
(in a sentence) parts which resemble other words. 

[Just as the abstracted parts of diphthongs like ai, au seem 
to resemble independent phonemes like a , 0 , i, u etc., in the same 
way, the abstracted parts of a sentence seem to resemble in¬ 
dependent words. But that is not true because these parts 
have no real existence.] 

12 . The words Vrsabha , Udaka and Tavaka 
have meaningless parts. Analysis through the 
method of agreement and difference is the basis of 
worldly usage. 



[Everywhere the parts which we artificially make in a word 
are meaningless. Ifwe make these parts and attribute meanings 
to them, it is only in an artificial sense. Words like Vrjabha, 
Udaka , Yavaka contain meaningless parts in the form of Rjabha, 
Uda and Yava respectively. Similarly, words like pacati> 
pacaka have meaningless parts in the form of stem, suffix etc. 
But the splitting of a word into stem and suffix and attributing 
meanings to them atleast serve the purpose of teaching the 
formation of words to the ignorant. But the artificial parts 
Rfabha, Uda and Yava in Vr$abha , Udaka and Ydvaka do not 
serve any such purpose.] 

13. The word has no division, how can the 
meaning have any? But the ignorant person sees 
division through artificial splitting. 

[The available Vrtti on the Ilnd Kanda of the Vakyapadiya 
begins from this verse and goes on till the end, with many gaps 
big and small, in the middle. On this verse, what it says can 
be summarised as follows—The sentence, the Vdkyaspho\a> 
is indivisible. Its meaning, namely, pratibha , is equally indi¬ 
visible. This has been stated also later on in verse 145. This 
artificial division of the sentence and the sentence-meaning is 
a means to the understanding of the indivisible sentence and 
the sentence-meaning. To understand the indivisible straight¬ 
way is as difficult as learning each word of a language sepa¬ 
rately. As the Vrtti puts it :— guruprakramd tvatra saqifrfta- 
rupasya pratipattir avibhagena, pratipadam pathavat = “The under¬ 
standing of the unified word without dividing it is a difficult 
process. It would be like learning each word of a language 
as a whole. 5 ’ A good student, however, knows that division is 
only a means to an end and that indivisibility is the truth. 
In any case, doubts sometimes arise as to how the division 
should be effected. For instance, in the case of words haridru 
and dustara. Even when it is effected, it is only a means to an 
end. The word is really indivisible.] 

14. Just as, in the compound, brahmanakam - 
bala, the term brahmana has no meaning, in the same 
way, in a sentence, words like Devadatta are without 
a meaning. 

II. 15-16 


[In the compound word brdhmanakambala , the element 
brahmana has no meaning. Similarly, in the sentence : Deva- 
datta gam abhydja dandena , the words Devadatta etc. have no 
separate meaning. Individual words are, therefore, meaning¬ 
less. The truth is that the sentence is anindivisible self-suffi¬ 
cient unit, quite different from the individual words. Simi¬ 
larly, the sentence-meaning is an indivisible, self-sufficient 
unit. When the compound brdhmanakambala is uttered in 
isolation or after some words or sentences, even though the 
word brahmana is heard and some meaning understood from it, it 
is known to have no meaning. Similarly, even when the word 
is not uttered in isolation or after other words and sentences, 
parts of sentences which resemble individual words do not 
really exist and have, therefore, no meaning.] 

15 . The general meaning, having disappeared, 
cannot lead to the particular. How can what has 
already been conveyed be abandoned and where 
can what is already gone rest ? * 

[The M imams aka view is that it is the individual words 
themselves which constitute the sentence and not anything 
beyond them. This is not sound. Words are uttered in a 
sequence. When the second word is uttered, the first one has 
already vanished. So its meaning which is of a general nature 
cannot become particularised in association with the meanings 
of the later words because it is not there at all. Even if it has 
an existence in memory, how can the general meaning which 
was first conveyed be abandoned ? To do that would go 
against the eternal relationship between the word and the 
meaning, which the Mimavisaka would not accept. In any 
case, if the word gives it up, where would it rest? 

The Vrtti points out that a word which, at the time of its 
utterance, conveys the general cannot denote the particular 
after its disappearance. Nor can the speaker mean to convey 
the general and the particular at the same time : na ca samanya- 
vi§e$ayor vivak$a yugapat sambhavatiJ] 

16. If the sentence-meaning is not derived 
from the words (but from their interconnection), 
the meaning of the word would also be in the same 

8 VAKYAPADIYAM of bhartrhari 

position. Thus the relation between the word 
and the meaning would be destroyed. 

[The criticism contained in the above verse is answered 
as follows— After all, a particularised meaning is understood 
from the sentence. How to account for that? The Mlmarn- 
saka view does not explain it either. If the particularised 
meaning does not come from the words which cease to exist 
after conveying the general meaning, it means that it does 
not come from words. It is aSabda. The same thing can be 
said about the meaning of the individual word. Because, the 
phonemes are uttered in a sequence and they disappear. So 
they do not co-exist and cannot constitute a word. So there 
cannot be a word-meaning. Thus, the relation of expression 
and expressed between word and meaning, accepted by the 
Mimamsaka> would cease to be. 

The Vrtti argues differently to reach the same conclusion. 

17. According to some, words expressive of 
the particular ( visesa ), resembling those which are 
expressive of the general, become clear to listeners 
when they are (later) connected with the other 
words in the sentence. 

[When a sentence is uttered, it is done in order to express a 
particularised meaning. So the very first word, even when 
it seems to be the same as in some other sentence, expresses, 
not a general meaning, but a particularised one. This becomes 
clearer when each succeeding word is heard. So the meaning 
of the sentence is the meaning of each word as connected with 
the meanings of the other words. No word conveys an un¬ 
connected meaning. It conveys an idea with its connection 
with other words implied. This is an explanation of the two 
views regarding a sentence, referred to as padam adyam and 
prthak sarvapadam sakanksam. 

The Vrtti explains this verse in one long and rather obscure 
sentence. It seems to say this : The very first word of a sen¬ 
tence denotes its meaning as qualified and delimited by the 
meanings of the other words of the sentence which are yet to 
come and when they do come, they do not say anything new, 

II. 18-19 


they only make clear and patent what is already latent in the 
meaning of the first word. The very first word is as good as 
the sentence.] 

18. According to them, the whole of the sen¬ 
tence-meaning is concentrated in each word. 
Hearers understand the meaning all the better 
when all the expressions are uttered. 

[Nor can one object that if the very first word conveys the 
meaning of the whole sentence, the other words would be 
useless. The first word conveys the sentence-meaning only 
vaguely. The others make it clearer. But this view is not 
accepted because the utterance of the other words would be 
either for the sake of restriction or restatement. Neither is 
necessary if the very first word or any word in the sentence 
can convey the meaning of the whole sentence. As a matter 
of fact, neither the first word ( padam adyani) nor any word 
•separately ( prthak sarvapadam) can convey the meaning of the 
whole sentence. If the other words are uttered at all, no matter 
for what purpose, the anvitdbhidhdna view falls. 

What the Vrtti says on this verse amounts to this: 
According to those who accept the use of many words to ex¬ 
press one meaning, all the particulars, the whole of the indivisible 
sentence-meaning is concentrated on each word and on each 
phoneme. As the Vrtti puts it— te$dm evamupagrhitascirvavi - 
Jesa ekasminnarthe bahuSabddnabhyupagacchatdm avikalpah krtsno 
vakyarthah pratipadam prativarnam va samapyate. Thus, when 
all the expressions are actually uttered, the sentence-meaning 
becomes clear to the hearer and not by its mere existence be¬ 
fore all the words are uttered.] 

19. The indistinct word which is uttered in 
silence appears to have sequence. The intelligence 
perceives it as having extension (that is, sequence) 
though it is really without any. 

[The sentence is really indivisible. But it is manifested 
by the sounds and it seems to have divisions like the phonemes 
and individual words. This indivisible Word has an inner 
and outer aspect. As something existing within the speaker, 



it is inner. As it is externalised by utterance, it is outer also. 
It is indivisible and eternal. It is this which is expressive of 
meaning. All distinctions such as spoken to oneself, spoken 
slowly, spoken quietly etc., are based on the sounds which mani¬ 
fest it and not its own. When one utters something well within, 
to be heard by oneself only, it is called upamfu. When no 
utterance takes place, that is, when the prana plays no part 
but the word is thought of in one’s mind and has assumed 
divisions and sequence, it is called paramopainhi. As the Vrtti 
puts it— Tatra pranavrttyanugrahe satyevayatra Sabdarupam parair 
asainvedyam bhavati tadupainSu. Antarena ta pranavrtiyann - 
graham yatra kevalam eva buddhau samavi$tarupo buddhyupadana. 
eva sabdatma tat paramopamhi . When the activity of the mind 
is suspended and the divisions and sequence are not there but 
due to their impressions they are superimposed, it is pratisam - 
hrtakramam. ] 

The next two verses are meant to elucidate the view that 
the sentence is jatih samghatavartini. 

20 . Even though it is particular movement 
which is made the specific action is not cognised 
but when it is repeated, the universal of actions 
like turning is manifested. 

21 . In the same way in the case of (the sphota 
of phonemes, words and sentences which may be 
different from another, the manifesting sounds 
appear to be the same. 

[A movement like turning consists of a series of 
momentary movements. Each movement consists in coming 
into contact with a point of space and then disappearing. It 
is then replaced by the next movement which does the same. 
As these point-movements are momentary, they cannot co¬ 
exist and cannot form a whole of which they would be parts 
and in which the universal of the movement called turning 
can inhere. Such a universal, different from the wider universal 
caUed ‘action’ is accepted and if it does exist, it would inhere 
Tn the momentary movements relating to turning. Even though 
this universal exists in each point-movement, the latter is not 

II. 22-24 


capable of producing the cognition of the universal because of 
too much resemblance between the movements of turning and 
the movements of another movement like, say, lifting. The 
movements of each movement are the result of a special effort 
to produce that movement and so they are the substrata of the 
universal of that movement. But that universal cannot be 
cognised until a series of point-movements has been cognised. 
It is not enough to cognise one or two such point-movements. 
The universal really inheres in each movement and it is also cog¬ 
nised. But the cognition is too vague and unfit for verbal 
usage. After a series of movements is cognised, the direction of 
the movement becomes clear and so the cognition of the univer¬ 
sal, inherent in each movement also becomes clear. In other 
words, it becomes quite clear that the movement is turning 
and not, say, lifting. The process is similar in the perception 
of the Word (Sentence.)] 

22 . How can there be really priority or posterio- 
rity in eternal things ? It is the power of the One 
that it so appears. 

[The Sentence is One and indivisible without any sequence 
involving priority and posteriority. But it has within it the 
power to appear as many, as having sequence. It is really 
the sounds which manifest it that have sequence but the vakya - 
sphota itself appears to have it.] 

23. Just as the cognitions ‘slow 5 and ‘quick 5 
appear to have temporal difference even though 
they have none, in the same way do long and short 
vowels appear to have temporal difference. 

[The cognitions‘slow 5 and ‘quick 5 have, as cognitions, no 
difference in time, but appear to have it because of the differ¬ 
ence in their object. In the same way, the sphota has no dis¬ 
tinction such as slow and quick but appears to have it because 
of the difference in the manifesting sounds.] 

24. Time, which is eternal, cannot be differ¬ 
entiated through divisions belonging to something 
else. (According to the view that there is no eternal 



entity called Time) the phenomena, being of a 
transient nature, cannot exist at the same time. 
How then can there be sequence ? 

[The latter half of the stanza is an answer to an objector 
who might say that he does not believe in Time as an entity 
separate from the phenomena. Time is nothing more than 
the phenomena themselves following one another, in succession. 
It is on the basis of these phenomena that we speak about 
sequence and simultaneity and not on the basis of a separate 
entity called Time. 

The main idea in the Vrtti on this verse may be expressed 
as follows—If Time is eternal, One and indivisible, it cannot 
have divisions on the basis of something external like action. 
Secondly/ Time conditioned by some action, is a measure. 
Therefore, there cannot be something else to measure and 
divide it. If everything is eternal, it cannot bring about 
division in Time. If everything is momentary, there cannot be 
any sequence between any two non-simultaneous moments 
and, therefore, sequence cannot be attributed to time on the 
basis of the sequence of the momentary objects. As there is 
no sequence in the things or in Time, to see temporal difference 
in long, short and prolated which are temporally identical is 
not right. 

The words of the Vrtti say something like what I have said 
above but I do not claim that the idea is very clear to me.] 

The idea that the notion which arises on the basis of these 
phenomena would explain Time is now answered. 

25. The cognition which is produced by them 
is One and indivisible but through its own power, 
it seems to be differentiated and assumes sequence. 

[Though bhagavartini is found in some editions and manu¬ 
scripts, bhagavarjita is a better reading and is confirmed by 
Punyaraja’s nirvibhdga and the Vrtti 9 s bhagavarjita . The gist 
of the Vrtti seems to be this:—Cognition is One because it 
is essentially in the nature of illumination, (sa tveka bhagavarjita 
buddhih prahasasvarupatvat...). Therefore its homogeneous and 
heterogeneous divisions are one with it. Being of the nature 
of illumination, cognition is partless and sequenceless, but it 

II. 26-27 


displays the sequence of the objects now mixed up with it 
within itself. Strictly, it (the sequence) should not be so dis¬ 
played but the cognition displays it without dissociating it 
from the objects. Such indeed is the way in which cognition 
displays sequence.] 

According to Punyaraja, the next verse is meant to refute 
the following view:—Even though cognition is one and indi¬ 
visible, it appears to have parts and sequence because of the 
sequence of the objects which figure in it. The cognition 
appears to have the sequence of the objects. This is due to an 
immemorial and eternal predisposition of cognitions. This 
appearance of sequence in the cognition is not contradicted. 
It is of the nature of cognition itself. A cognition displaying 
sequence leaves a similar impression ( samskara ) so that the next 
cognition also displays sequence. That is why in our cognition 
of short, long and protracted vowels or of words and sentences, 
temporal distinctions figure though these are eternal and have, 
therefore, no temporal distinctions. 

26. It is not possible to explain as identical 
with the intellect or as different from it the seed 
which is sown in it by the display of sequence (in 
the cognition). 

[The impression of sequence which is supposed to be left 
by the cognition in the intellect is either identical with it or 
different from it. If it is different from it, it cannot do any 
service to it. It cannot bring about temporal distinction in 
short, long etc. If it is identical with it, then intellect or cogni¬ 
tion would remain one and indivisible and there would be no 
sequence at all. Therefore, it is futile to try to explain sequence 
in cognition through the impressions of the sequence of objects 
which figure in the cognition.] 

Indivisibility is now set forth in a different manner. 

2 /. This indivisible (Sentence) is understood 
as having the capacity for sequence (or division) 
at the time of the artificial separation ( bhavana - 
samaye). Hence its meaning, though devoid of all 
division is perceived as having division. 



[In the Brahmakanda, the nature of sphoia has beendeter- 
mined. It is the sentence as sphota which is expressive. At 
the time of analysis, the sentence-meaning is understood as 
having sequence and division, but it really has none. 

The Vrtti gives the analogy of Dik to explain how the One 
appears as many. Dik (Direction, Space) is one but when 
we use expressions like urddhvam, adhah , tiryak , we are speaking 
about it as though it had divisions, knowing that it has none: 
urdhvam adhas tiryag ityekatve’pi disamyathayatham vinivif tab/ia- 
vanavisefdh sarvapraninah= All beings have an inner disposition 
to divide Dik (Direction, Space) which, though one, is regard¬ 
ed by them as being above, below, across and so on.] 

The view that the sentence is nothing more than a collec¬ 
tion of words and not an entity over and above them and that 
the word is nothing more than a collection of phonemes and 
not an entity over and above them is now criticized. 

28-29. If the words in the sentence are those 
very ones (which are found singly outside the sen¬ 
tence) and if the phonemes in a word are those very 
ones (which are found singly outside the word) 
then there would be divisions like atoms in the 
phonemes also. 

As the parts cannot combine (because of their non-exis¬ 
tence) there would be neither phoneme nor word. As they 
cannot be determined what other thing could be determined 
as the expressive element ? 

The Vrtti argues as follows— 

[If it is held that the words in a sentence are the very ones 
which are found separately and if the phonemes which are 
found separately are the very ones found collected in a word 
and if it is held that there is no essence of the sentence or of the 
word apart from the phonemes, then there would be atom¬ 
like divisions within the phonemes also by going to the extreme 
limit of division. These divisions would have sequence and 
would never be simultaneous and would, therefore, never come 
into mutual contact. Thus there would ensue nothing called 
phoneme or the word. The final divisions being indefinable. 



they would not enter into usage. Therefore what would be the 
verbal unit which would be clearly defined as ‘this 5 ? There 
cannot be any usage with a verbal entity the divisions 
of which arc indefinable and which are not collected together 
in something different from knowledge. The idea is that the 
sentence is a collection of words and if the word is only a 
collection of phonemes, one can divide the phonemes also. 
Ultimately, there would be no unit at all which would be looked 
upon as the expressive element. Therefore, the sentence is an 
entity over and above the phonemes and words. This entity 
is described as something outside the mind. But the Scistra 
is of the view that the sentence is buddhyanusamhrtih .] 

The view that the sentence is buddhyanusamhrtih , mentioned 
in the very first verse of this Kanda is now explained. 

30. Others declare the Word (that is, the Sen¬ 
tence) to be the word-Principle which is within 
and One and is manifested by the sounds. It is 
that unity which appears in the manifested sentence. 

[Others emphasize the inner aspect of the word. Its ex¬ 
ternal aspect as jati or vyakti has already been considered before. 
The chief characteristic of this inner aspect is that it is one and 
indivisible and without any inner sequence. It is an inner 
entity, consisting essentially of cognition or consciousness and 
an amalgam of word and meaning. When manifested by the 
sounds, wrongly thought of as its parts, itbecomesthe external 
sentence. The Vrtti points out that this external sentence is 
like the written symbols ( aksaracihnavat) which are mistaken to 
be the word. The external sentence is only a symbol of the real 
sentence which is within and is an indivisible unity.] 

It is now stated that not only the sentence but the sentence 
meaning also is indivisible and manifested by the word meanings. 

31. According to them, the inner meaning, 
(that is, the Sentence-Meaning) is manifested by 
parts of it. Word and Meaning (that is, Sentence 
and Sentence-meaning) are inseparable ( aprthak - 
sthitau ) divisions of one Inner Principle. 



[Just as the Sentence is an inner entity, so is the Sentence¬ 
meaning. They are identical with each other and with the 
One Inner Principle but, externally, they appear to be differ¬ 
ent from each other. Pur.yaraja does not say anything more. 

The Vrtti interprets this verse in a deeper manner. From 
its terse and rather obscure language, one understands some¬ 
thinglike this: — 

It is well-known that the Word-Principle is mainly the 
indivisible inner entity and that it is grasped through its inde¬ 
finable and unreal parts. Similarly, the meanings reflected 
in the intellect are experienced as identical with the external 
objects. This is according to the view that the external object 
is transient. According to the view that it is eternal, 
it manifests itself according to the power of sequence of the 
intellect {Kramasaktirupanirbhdsama.traya) . An external object 
is not fit for practical purposive usage without the intellect 
with which it is wrongly identified. All worldly usage is done with 
objects which have been grasped by the intellect. Thus both 
the word and the object are in the intellect. Others, on the 
other hand, declare that as the intellect in which the object 
is reflected inheres in the Self, the latter which is essentially 
consciousness, assumes the form of the intellect which inheres in 
itand this explains the experience ofthe individual {piiru$arth asya 
prasiddhim). When the intellect which is transparent assumes the 
character of consciousness and of the object through their 
reflections, the powers ofthe enjoyed and of the enjoyer ( i/zogjrt- 
Sdkti and bhoktrsakti ) quite distinct from each other and not at 
all mixed up with each other, seem to attain the state of non¬ 
distinction in the intellect, which has apparently assumed the 
nature of consciousness and of the object and then distinctive 
worldly usage becomes possible. The power to experience is 
unchangeable and is not reflected anywhere but it appears 
to be reflected in the intellect. It is through the imitative 
transformation of the intellect in which the consciousness is 
reflected that one speaks ofthe operation of specific knowledge. 
The transformation of the intellect is said to be imitative {anu- 
karamatrd) because, on the onehand, it imitates the objectwhose 
form it takes and, on the other, consciousness is reflected in it. 
In the background of all statements, there are three views :— 

n. 32 


(1) The external word is transitory and is an unreal manifes¬ 
tation of the eternal inner word. (2) According to satkarya- 
vada, all objects are eternal and they are experienced when 
manifested. (3) When the object is reflected in the transpar¬ 
ent intellect ( Buddhi ), it is experienced. According to all 
the views, the inner word is One, with all multiplicity resolved. 
From it, meanings which are also inner and undivided, with 
all multiplicity resolved, are understood without their original 
division being abandoned. What then remains is the single 
entity Intellect in which multiplicity has been resolved, in which 
the distinct powers of being the conveyor and the conveyed are 
not separated (pratipadakapratipattavyaSaktyor avibhagcna) even 
though their distinction is not abandoned. In the view that 
both the word and the meaning are eternal, neither is adven¬ 
titious, subject to increase and decrease and to change.] 

If the word and the meaning are identical, how are they 
said to stand in the relation of expression and expressed to each 
other ? 

32. Of that Word-Principle which is within, 
there is always this character of being the Illumina¬ 
tor as well as the Illuminated, the Cause as well as 
the Effect. 

[The idea is that the Inner Word-Principle has got both 
powers : that of being prakcUaka and that of being prakdsya, 
that of being the Cause and that of being the Effect. The One 
inner Reality contains the seeds of all manifestations, as 
already stated in Vak.1.4. 

In explaining this verse, the Vrtti emphasizes the nature of 
the Sabdatattva within. The inner word is endowed with two 
powers : the power of being the Illuminator and that of being 
the Illuminated. Though the latter is not separate from the 
former, it appears to be so. As they are mutually dependent, 
there is the relation of causality between them. As the Illu¬ 
minator, it is the cause and as the Illuminated, it is the effect. 
The former power is constant and the latter occasional ( anapayini 
and apayini). The Word-Principle is not associated with any 
particular place, but, because of its association with the intellect, 
it is spoken of as the inner word. Nor is it associated with 



time, divisible into past, present and future. That is why the 
word sarvada is used in the verse. The Word-Principle, being 
identical with Consciousness, is beyond existence and non¬ 
existence (bhdvdbhdvavibhdgabhdvdt). The expressions past, 
present and future.can be used only in regard to bhava or 
abhava .] 

What other power the Word-Principle has is now stated. 

33. It is that (the Word-Principle) whicn has 
the two powers of existence and non-existence which 
have no sequence but appear to have it and are the 
basis of all worldly usage. 

[The inner Word-Principle has the two powers of existence 
and non-existence. These two powers cannot have any sequence 
because of the eternality of their substratum, but they appear to 
have it. The indivisible sentence appears to have the sequence 
of the phonemes and individual words. Their existence and 
non-existence seem to have sequence. All manifestation involv- 
ing multiplicity and sequence is due to the powers of the Word- 
Principle and not merely its manifestation as the Illuminator 
and the Illuminated]. 

The view that the individual word has a reality is again 

34. If the mere understanding of the word¬ 
meaning is taken as proof of the reality of the word 
and its meaning, as it is abandoned when it is mixed 
up with the meaning (of the following words) it 
would become useless. 

[If the mere fact that the individual word and its meaning 
are cognised, are taken as proof that they have reality, there 
would be a difficulty. Each subsequent word of a sentence 
conveys its meaning as mixed up with that of the preceding 
word or words and so the meanings conveyed by the preceding 
word or words are abandoned as such. Thus the utterance 
of the preceding words becomes useless. Therefore, when 
the sentence-meaning is ultimately understood, it would not be 
due to the useless preceding words. The understanding of the 

II. 35-36 


individual word and its meaning were only temporary and, 
therefore, unreal phases.] 

It is now stated that, as the individual word is something 
indefinite, its reality should not be accepted. 

35. (ab). The word raj an conveys the idea of 
‘king’ through its different forms. 

[The word rajan does not occur in speech in a particular 
form but in its different inflexional forms like raja, rdjdnam , 
rajna , rajiie , rdjnah , rdjni (or rajani). If the word had a reality, 
its form would not thus change.] 

35. (cd). In a complex formation, a word (a 
noun) resembling a verb could mean something 
different (from what it ought to mean). 

[The compound word rajapurusa could be understood to 
mean : Shine ! O Man ! and yet.that is not its meaning. 
Which shows that what looks like an individual word has no 

It is not only independent individual words which have 
no definite fixed form but as parts of compound words also they 
have no fixed form. 

36. Just as the compound word asvakarna con¬ 
veys a special meaning without the part asva con¬ 
veying any meaning of its own, so do other compound 

[We understand from the word asvakarna a particular class 
of objects, having a particular universal inhering in it. The 
meaning of the part afva does not contribute to it. The mean¬ 
ing comes from the word as a whole. It is better to be¬ 
lieve that the same thing happens in all compound words. 
If the meaning of asva plays no part in the meaning of the whole, 
it is wrong to think that there is a part called a$va in the com¬ 
pound at all. Similarly, there is no part called karna in it. 

Here the Vrtti points out that, for the purpose of grammati¬ 
cal derivation, one may explain the meaning of the compound 
asvakarna as asvasya Karnaviva parnau yasya = a tree of which 
the leaves are like the ears of a horse. But this is only artha - 


VAKYAPADIYAM of bhartrhari 

pradarsika = ‘showing of an alleged meaning 5 and should not 
be taken seriously. It does not prove the reality of parts.] 

If all compound words are in the same position, what 
would be the basis of the distinction between rudhi *= ‘conven¬ 
tional 5 words and arudhi= c non-conventional 5 words ? 

37. While resemblance is postulated (bet¬ 
ween the meaning of the compound word and the 
analytical sentence which explains it) in the case of 
some compounds where the analysis conveys a 
totally different meaning, the Science of Grammar 
itself looks upon them as conventional words. 

[A difference is pointed out between compound words 
like rajapurusa and tailapayika. The meaning conveyed by the 
former and the meaning obtained from its analysis are very 
similar. So it is not looked upon as a mere rudhi word. The 
meaning conveyed by the second compound is that of cockroach, 
a kind of insect. The meaning obtained by analysis is : 
“that which drinks oil 55 which is quite different from what the 
whole compound means. That is why it is a rudhi word. 
In regard to such words, grammar teaches some special opera¬ 
tions as in P. 6.2.76, 77, 146 ; P. 6.3.53 etc. This is pointed 
out by the Vrtli.] 

38. Means ( upayah ) are so called because they 
are first accepted ( upadaya) and then abandoned 
[heyah). And no fixity necessarily exists in regard 

to the means. 

[The Vrtti points out that when correct forms are being 
taught in the Science of Grammar, as in the case of compounds, 
secondary derivatives, denominative verbs and ekase$a , indivi¬ 
dual words are brought in for making analytical explanatory 
sentences. When the purpose of teaching their correct form 
is achieved, these individual words are abandoned. Similarly, 
some rules relating to accent (such as P. 8.2.4, 5, 6; P. 6.2.2,65) 
teach correct forms of words on the assumption that there are 
individual words in compounds or sentences, that they have 
their own meaning and that they come in a certain order. 

II. 39-40 


All this has to be understood in order to grasp the correctness 
of the forms in question. Once that is done, the rest is for¬ 

39. One understands the facts in some way or 
another. Whether united or separated, the divisions 
are based on the sentence. 

[One can understand the facts of the Sanskrit language 
either through Panini’s grammar or through some other. All 
of them have to analyse the sentence and sentence-meaning to 
do their work. The latter are, therefore, the real entities and 
not what is obtained by analysis. 

The Vrtti gives examples of the different ways in which the 
same word is derived, by different grammarians. The word 
srotriya is one of them. According to some, it is a word 
expressive of the meaning of the whole sentence chando’dhite= 
£ he studies the Vedas’. Others say that the word chandas 
becomes srotra to which the suffix ghan is added in the sense of 
tad adhite (P. 4.2.59.) or in the sense of action done by the ears. 
Similarly grammarians differ in their derivation of words like 
ustragoyugae and asvasadgavae . The Vrtti also gives two expla¬ 
nations of the second half of the verse. The difference is chiefly 
in regard to the meaning of the words samsrstall and vibhaktdh . 
In the first explanation, the former means ‘united’, that is, one 
word and its meaning united with another to form a compound. 
The second word means ‘separated’, that is, each word stand¬ 
ing separately in the sentence. Whether a word combines with 
another to form a samasa or not, its meaning is determined 
by the sentence as a whole. In the second explanation, the two 
words mean ‘included’ and ‘excluded’ respectively.] 

It is now stated that the sentence-meaning can sometimes 
be understood even from a single phoneme. 

40. Sometimes, when the relation of identity 
in the form c it is this 5 is understood, then even a 
single phoneme can be expressive of the meaning of 
the whole sentence. 

[When the identity of the sentence and the sentence-mean¬ 
ing is already in the mind of the speaker, then it is not necessary 



that the sentence-meaning should be understood after the word- 
meanings have been understood in the sequence in which the 
words have been uttered. A mere phoneme without any inner 
sequence may sometimes be able to convey it. Punyaraja does 
not give any example. The Vrtti does and ends by referring 
to the way in which the sages, those who are experts in Vagyoga, 
grasp the integrated Word-Principle within.] 

Now the view of the sentence referred to by the word 
sanghata in the very first verse is being explained. 

41. It has been declared that a word, as part 
of a sentence, expresses the same extent of meaning 
as it does when it is in isolation. 

[Punyaraja points out that this samghata view is held by 
the Abhihitanvayavadin. The Vrtti illustrates by saying that the 
isolated word Vrkfa conveys a particular universal, namely 
tree-ness. It does the same in the following sentences: Vrkpo’sti, 
Vrfoo nasti Vrk$a$cchinnah. Because of its connection in these 
sentences with existence, non-existence and cutting, it does 
not become associated with another universal. In any case, 
these actions are not connected with a universal, but with the 
individual in which the universal inheres.] 

42. Whatever extra meaning is understood 
when the words (in a sentence) are connected to¬ 
gether is the meaning of the sentence and it rests 
on many words. 

[According to this view, the sentence meaning is the inter¬ 
connection ( samsarga ) of the meanings of the individual words. 

The Vrtti points out that the extra meaning which comes 
when the words are joined together does not play any partin 
determining the form of the individual words. They get their 
form on the basis of the meaning or meanings which they them¬ 
selves express. 

The words of somebody, referred to as tatrabhavan, are 
quoted in the Vrtti. In the writings of Abhinavagupta, tatra¬ 
bhavan always stands for Bhartrhari. If that is the case here, 
it means that the Vrtti is not by Bhartrhari. But the title can 
very well stand for somebody else. Who is the question?] 

II. 43-45 


Another view on the subject is now given. 

43. Even though it rests on many words, it is 
found in full in each like the universal, or, like 
number, it is said to rest, by others, on the whole 
collection of words also. 

[The universal is supposed to exist in every individual in 
full. Number is supposed to exist in all the units together. 
Similarly, the view that the sentence-meaning rests on many 
words is understood by some to mean that it exists in each 
part of it in full like the universal and by others to mean that 
it exists in the totality of words also. Those who hold the 
latter view go by the principle that what does not exist in the 
part cannot exist in the totality. As the Ambakartrl puts it: 
na hyavayavaparyaptasya samudaye paryaptir asti. Both Punya- 
raja and the Vrtti speak about these two views. 

The samghata view is now explained from the anvitabhidhana 
point of view: 

44-45. Others declare the meaning of the word 
to be so general as is adaptable to all the particulars. 
When connected with the meaning of the other 
words, it assumes the form of the particular. 

Contact (with the meaning of the other words) removes 
the vagueness which arises when it requires particularisation 
by fixing it to one particular. 

[It was stated in verses 41 and 42 that a word in a sentence 
gives the same meaning as in isolation. When the meanings 
of the words are connected together, a qualified meaning 
emerges and that is the meaning of the sentence as a whole. 
In verse 43, it was stated that the meaning of the whole is con¬ 
tained in a single word according to some and in the totality 
of words also according to others. In 44 and 45, it is stated 
that the individual word conveys a general meaning which is 
potentially capable of being connected with the meanings of 
other words and when it is actually connected with the other 
words, it conveys a meaning actually connected with particular 
meanings of other words. The general meaning and the parti- 


VAKYAPADIYAM of bhartrhari 

cular meaning are those of the individual word and not of the 
sentence and is not conveyed by dkdiik$d,yogyatd and sannidhii] 

46. The connection is to be inferred from its 
effect; it has. no particular form. That is why 
some declare it to be absolutely non-existent (as 
distinct from the meanings of the individual words.) 

[The effect referred to is the fact of the meaning of the 
individual word referring to the particular and not merely to 
the general. The connection inferred has no concrete shape or 
form. That is why it is said to be asattuabhfita.'] 

The same view is further explained. 

47. An accessory (sadhana) necessarily pre¬ 
supposes an action to be accomplished ( sadhya ) 
and an action is necessarily connected with acces¬ 
sories. This necessity which is present becomes 
evident in the presence (of the meaning of other 

[Action and accessory presuppose each other. It is only 
when the words expressive of them enter into the sentence that 
what is presupposed becomes manifest. So the sentence is 
nothing more than the individual words presupposing one 
another and the sentence-meaning is nothing more than the 
word-meanings presupposing one another. 

The Vrtti points out that this mutual requirement of action 
and accessory is one and indivisible. It is not felt in the absence 
of the other. It becomes manifest in the mere presence of 
the word expressive of the other and not because of the help 
of any other factor. As the Vrtti puts it : 

nityaniyatatvacca niyamo'pyatra vidheyah , sanstu niyamah 
prayoge sannidhimatrena padantardnam puru$am prati prakasate , 

If the mutual requirement of the meanings of the words 
of a sentence is equal, how to decide which is primary and which 
is secondary. 

48. In it, (that is, the sentence), the noun, 
being secondary (to action) requires the verb. The 

II. 49-51 


verb, being expressive of something to be accom¬ 
plished, requires (the words expressive of) the means. 

The meanings of the individual words of a sentence require 
one another. Some looked upon this mutual requirement 
as a property of the meanings while others looked upon it as 
property of the listener. The Vrtti refers to this difference of 

view. It says that the listener understands mutual requirement 
(;vyapekfa) according to what the words convey, whether it 
actually exists or not in the meanings : arthe$u satim asatim va 
sabdavrllyanukarena purufo vyapekfam samihate .] 

The view that the sentence is nothing more than the 
sequence of the words is now explained. 

49 . The particularisations which already exist 
in the word-meanings are understood from their 
sequence and there is no expressive sentence beyond 

[If the sentence is nothing more than the sequence, it is 
not anything verbal, it is not something which can be heard. 
As the Vrtti puts it : naparam vakyam nama kiiicicchabdarupam 
abhidhayakam vidyate .] 

The same view is further explained. 

50. The mere sequence of the words being thus 
expressive, there is no other verbal element which is 
so. And sequence is a property of time. There 
is no separate entity called sentence. 

[What is denied here is the existence of some verbal element 
characterised by sequence, which could be called the sentence. 
The sentence is nothing more than the sequence itself. Se¬ 
quence which is a property of time is superimposed on the words. 
What is called sentence is a mere name without any reality 
behind it. As the Vrtti puts it— tena vakyam ityavastukameve- 
damabhildpama.tram, padamevarthavaditi .] 

It is now stated that the same is not true of the sequence 
of phonemes. 

51 . The particularisations which already exist 
in the meanings of individual words but are not 


vakyapad!yam of bhartrhari 

evident become manifested in the presence of the 
other words. But such a (meaningful) sequence 
does not exist in the phonemes. 

This point is now concluded. 

52. The names Word and Sentence may be 
applied to the sequences of the phonemes and the 
words respectively but they are not the expressive 
element ( sabdatva ). 

[The phoneme and the word are audible but mere audi¬ 
bility does not entitle them to be called sabda. For that, they 
must convey the meaning, they must be vacaka. They are not. 
Only sequence is so.] 

53. Even though the fact of being word ( sabda , 
something audible) is common (to both the phoneme 
and the word), the meaning is understood from the 
word (conceived as the sequence of phonemes) 
but not from each phoneme. Hence the meaning 
belongs to the word. 

[As audible entities, the phoneme and the word are both 
sabda but the sentence-meaning is understood from the sequence 
of the words and not from the phonemes. So sequence 
is the sentence and interconnection of word-meanings is the 

The sanghata view is now restated. 

54. Just as phonemes, parts of a word are 
devoid of any meaning, so is the case with the sen¬ 

[All the phonemes put together may have a meaning, 
but^ach one, taken separately, has no meaning. Each indivi¬ 
dual word, by itself, cannot express the sentence-meaning which 
is in the nature of a connection between the meanings of indivi¬ 
dual words. Only a collection of such words can do it and so 
that is.the sentence.] 

Now the sequence theory is restated. 



55. Individual words, meaningless being only 
a means to an end or having their own meaning 
only when uttered in a sequence, convey the sentence¬ 
meaning which is different from the meanings of 
individual words. 

[The meaning of a sentence is the meaning of the individual 
words in it as syntactically connected with one another. No 
single word can express this mutually connected meaning. 
There is not only temporal sequence between the words, but 
also syntactic connection.] 

The sphota view is now stated. 

56. The sentences, looked upon as collections 
of words, being eternal or when the sentence-uni¬ 
versal is postulated, it is the one sentence which 
expresses an inalienable meaning. 

[The sentence may be looked upon as one unit. As uttered 
by different persons, it is something different and yet it is recog¬ 
nised to be the same. The different sounds uttered by the 
different speakers ultimately manifest the same sentence. Or 
the sentence may be looked upon as a universal which inheres 
in the different utterances of it by different persons. In whatever 
way it is looked at, it is one and it expresses a meaning in 
the nature of intuition (pratibha) and is indivisible.] 

57. According to the upholders of the (indi¬ 
visible) sentence, unity precedes division which is 
fictitious. The upholders of the individual word, 
on the other hand, hold that the unities of the sen¬ 
tence are preceded by their divisions. 

The statement padaprakrtilj sainhita (Rk-pra.lisakhya.2. 1.) is 
now considered. 

58. The fact of the sainhita (the connected 
text) being the source of the individual words is 
explained by resorting to a different complex for- 



mation ( Vrtti ). padaprakrtih can be explained either 
thus : ‘the samhita is the source of the individual 
words 5 or thus : the samhita has the individual 
words as its source. 

[The compound word padaprakrtih is laken as a fhi- 
tatpurusa according to those who follow the akhandapaksa. 
The others take it as a bahuvrihi . If the Vedic sentences are 
indivisible and not composed by humans (apaurufeya), then the 
individual words obtained by analysis are of human origin 
( pauru$eya ). On the basis of this very statement, the other 
view, namely, that the individual words are real and that the 
sentence is a fiction can be justified. It is like this: The 
individual words are eternal and not formed by men. Their 
connection is man-made. Each word conveys its own mean¬ 
ing, plus its connection in general with the meanings of the other 
words. In the presence of the other words, this connection 
becomes specific and is realised. Thus, the sentence-mean¬ 
ing is conveyed by the words themselves and not by the 

The Vrtti also refers to the statement of the Btk-pratisakhya 
and says that some held the padapatha of the Vedas to be 
eternal and the connected text (samhitapa tha) to be of human 
origin while others held just the opposite view.] 

A statement of Patanjali is now explained. 

59. If the other tradition, namely, that of the 
separated words ( padapatha ) shows the way to the 
connected text and is eternal, how is the separated 
text to be constituted according to the rules ? 

The very fact that Patanjali says that the authors of the 
padapatha have to follow the rules in doing their work shows 
that, according to him, that text is man-made. His words 
are : na lakfanena padakara anuvartyah; padakarair ndma lak$anam 
aiiuvartyam== “Joules are not to be framed according to the text 
made by the authors of the padapatha ; it is the latter who have 
to follow the rules.” In other words, Patanjali is in favour of 
the akhandapaksa. See M.Bha. II. p. 85,1. 4.5. 

II. 60-62 


[The Vrtti points out that the word used in the world is 
eternal. It is for the rules to conform to the eternal word and 
not for the latter to conform to the rules : atah sastrena nityasya 
laksyasydnuvidhdnam kartavya , na til nityena lak$yena sastram 
anuvidheyam .] 

60. Just as the meaning of the word is not 
understood from each phoneme, in the same way, 
the meaning of the sentence is not understood from 
each word. 

[Here the Vrtti makes an interesting observation. It says 
that the cognition of the sentence meaning is self-luminous, 
besides illuminating the external object. Being self-luminous, 
the cognition of the sentence-meaning is authority in 
itself. The sentence-meaning which its cognition grasps is 
one and indivisible and being authority, it confirms the unity 
and indivisibility of the sentence-meaning.] 

To meet this, the upholder of the individual word says— 

61. Just as the meaning of the sentence is un¬ 
derstood when all the words are together, in the 
same way, the meaning of the word is understood 
when all the phonemes are together. 

[The idea here is that each phoneme does denote the mean¬ 
ing of the word and so does each word denote the meaning of 
the sentence, but only when the other phonemes and the other 
words are also uttered in the same context. So the analogy 
brought forward by the upholder of the sentence is not valid.] 

The point is now further elucidated. 

62. Just as a minute perceptible object, when 
associated with something else, is perceived with it, 
in the same way, a phoneme becomes expressive 
(of a meaning) when it is associated with other 

[What is meant here is that a phoneme really has a mean¬ 
ing but that is understood only when the other phonemes of the 
word are uttered. All of them express the meaning together, 


VAKYAPADIYAM of bhartrhari 

but each one has the potentiality to do so. The hearer cannot 
understand it unless the other phonemes are also uttered. 

63. Just as some meaning is understood when 
a word is uttered, in the same way, when the pho¬ 
nemes are near one another, the same meaning is 

[The point sought to be made here is this :—Just as, when, 
the phonemes are together, they become expressive by pooling 
their powers ( parasparasaktyaveiavasat ), so do individual words 
together convey the sentence-meaning. So there is no need to 
postulate a sentence apart from the phonemes and the words. 
The Vrtti points out that the upholder of the meaningfulness of 
the phoneme can argue exactly in the same way as the upholder 
of the individual word.] 

It is further pointed out that if the reality of the individual 
word is denied, it would not be possible to have substitutes if 
what is prescribed is not available. 

64. The purpose of the text (the word vri- 
liihhih ) is restriction to one possibility which occurs 
to one (in any case) due to the power of the verb. 
Therefore, if the accessory in general is set aside by 
a particular accessory, 

65. If mere substance which occurs to the mind 
through the power of the verb yajeta is set aside by 
the mention of the particular substance, there would 
be no substitute if rice ( Vrihi ) is not available. 

[If the verb yajeta in vrihibhir yajeta brings to the mind sub¬ 
stance in general and not the universal of any particular sub¬ 
stance, then, if rice, denoted by the word vrthibhih , is not avail¬ 
able, a similar substance like barley can be used. But if it is 
held that the mention of a particular substance sets aside sub¬ 
stance in general also, then there can be no substitute at all, 
because the substitute barley would come under substance in 
general which has been set aside.] 

II. 66-68 


Therefore, another view is adopted as follows— 

66 . Therefore, the word rice, bringing the 
additional idea of rice, has a positive meaning 
( praptyarthah ) and does not set aside substance in 
genei'al as there is no opposition between the two. 

[According to this view, vrihibhir yajela is a positive in¬ 
junction and not a restriction. It specifies and does not exclude 
other things. From yajela, substance in general is understood 
because without some substance, yaga cannot be performed. 
From vrihibhih, we further learn that the particular substance 
is rice. If rice is not available, another similar substance is 
not excluded.] 

67. When substance in general ( dravyaiva ) is 
taken away by it (from other substances) with 
which it co-exists, other substitutes are not seen 
there (that is, in the sacrifice) because of impossibi¬ 

68 . The verb does not bring to the mind all 
particular substances as it does substance in general. 
A word indeed does not express all the meanings 
which it has. 

[Here the Vrtti explains as follows— 

If the verb yajela brings substance in general to the 
mind and not a particular substance and so the word 
vrihibhih is a positive prescription of rice and not an 
exclusion of other substances, it means that a substance 
like barley has not been excluded. That beingso, why doesit not 
come optionally ? The answer is that prescriptive words are 
of two kinds : (1) Some have a restrictive effect. They pres¬ 
cribe something and in effect, exclude others (2) Others do not 
emphasize the restrictive side. Some scholars of Mimamsa say 
that, in some matters this becomes, a kind of restriction 
through impossibility ( asambhavaniyamah) when, through the 
mention of the word rice, the universal of rice ( vnhitva) be¬ 
comes an extra accessory of sacrifice, it is not possible for other 
universals like yavatva to become accessories though they have 


vakyapadIyam of ishartrhari 

not been openly set aside, because they cannot coexist with 
Vrihitva in the same thing. If they could become accessories 
at all, it would be because they come to the mind due to lack 
of contradiction. They would then be combined with other 
things and not adopted optionally. Only, that which is openly 
stated by the words can become an option. The verb does 
not convey a particular accessory but only substance in general. 
A word does not express all that actually exists.) 

Now an illustration is given. 

69. Just as qualities like white, though present 
(in the rice) are not meant to be conveyed (by 
the word vrihi= rice) in the same way, theuniversals 
of being particular substances which co-exist with 
substance in general (i dravyatva ) are not meant to 
be conveyed. 

[A word does not express as its meaning everything that 
exists in an object. An object, to be expressed by a word, 
depends upon the desire of the speaker to speak about it. Even 
existing objects do not exist as meanings of words. The 
speaker’s desire to speak about a thing depends upon the capacity 
of the form of the word to express it: rupasdmarthyanjbandhana 
as the Vrtti puts it. Due to the natural power of words, the 
verb cannot denote the colour which exists in the substance in 
general which it brings to the mind. Similarly, the verb 
cannot denote all the particular substances in which dravyatva 
co-exists with their particular universal.] 

The purpose of substitution is now stated. 

70. A substitute is taught in the absence (of 
what is prescribed) in order that there may be no 
omission of a compulsory ( nitya ) ritual or of an 
optional one which has already been begun. 

[What is pointed out here is that even in the case of an 
optional ritual, if what is prescribed is not available, it should 
be performed with a substitute. This is all the more so in the 
case of a compulsory ceremony, because its omission would 
result in sin. As the Vrtti puts it : Nityam yat karma tasyakarane 

II. 71-72 


Thus it has been shown how a substitute becomes possible 
by taking the view that the verb brings only the substance in 
general to the mind and that the word vrlhibhih is a positive 
injunction and not meant to exclude other substances. It is 
now stated that if the sentence is indivisible, a substitute would 
not be possible. 

71. For him who looks upon a sentence as 
expressive of a particularised action, when a substi¬ 
tute is used for a substance which is not available, 
there would result a different action altogether. 

[It is accepted by all that, while another substance may 
be substituted if the prescribed one is not available, the prescrib¬ 
ed ritual, an act, should not be substituted by another. But 
if the sentence is looked upon as expressive of an indivisible 
meaning, what looks like substitution of material would really 
be substitution of ritual which is not allowed. The sentence 
being indivisible, it does not teach two things. It does not say: 
(1) One should perform the ritual, (2) One should do it with 
rice. That would imply that the sentence has two parts which 
the doctrine of unity and indivisibility does not accept. So the 
sentence teaches only one thing, a ritual to be done with a 
particular material. When that is not available and the ritual 
is done with some other material, it would not be a mere sub¬ 
stitution of material, but of the ritual itself. But that is not 
allowed. When the Veda enjoins one ritual, it would be wrong 
to perform another.] 

It is now stated that, if the individual word does not exist, 
one cannot explain why people sometimes enquire about the 
meaning of a particular word. 

72. When the meaning of the known words (in 
a sentence) has been understood, why does one 
enquire about the meaning of an unknown word 
like pika ? 

Punyaraja and the Vrtti explain this karikd in the same way, 
except that the latter gives two examples instead of one. It is 
well-known that listeners sometimes fail to understand the 
meaning of just one word in a sentence and enquire about it. 


VAKYAPADIYAM of bhartriiari 

In vandt pika aniyatam, the word pika may be obscure to 
somebody, who would therefore enquire about its meaning. 
In the sentence vdrangi jarjard vrfalaya diyatdm= let the torn 
z/dra 7 i£i'=‘turmeric-coloured dress’ be given to the jur/ra, some¬ 
body may not understand the meaning of Vdrangi and may just 
enquire about that. This shows that the enquirer has the 
consciousness that the individual word has a meaning. No¬ 
body makes a similar enquiry -about the meanings of a or 
Kin the words Vrjabha and kandira after having understood the 
meaning of the remaining portions of the two words, r^abha 
and andira, showing that people have no consciousness that 
phonemes have a meaning.] 

73 -74. When what is implied is mentioned for 
the sake of clarity, it is a case of direct statement 
(sruti) and it sets aside indication ( lihga ) and juxta¬ 
position ( vakya ). 

A quality like white, when not implied, is 
understood through proximity and is therefore 
conveyed by a special effort and is different from 
direct statement. 

[If the individual word has no reality, the principle that 
when there is opposition between direct statement and juxta¬ 
position, the former prevails would not work. Six principles 
are laid down by Jaimini in order to determine the meanings 
of doubtful sentences. Of them, each preceding one is stronger 
than the following one. In the sentence : svetam chagam alabhe- 
t a =‘ 0 ne should sacrifice a white goat’ the connection of the 
act of sacrificing with the goat is conveyed by direct statement 
(sruti), namely, the second-case-affix. The quality white 
becomes connected with the act because the word expressive 
of it is in apposition to the word expressive of the goat. In 
other words, through juxtaposition. If both the words— 
chagam and Svetam —are connected with the verb at the same 
time because of the second case-affix, then the connection in 
both cases would be through sruti, with the result that if either 
is not available, there can be a substitute.] 

75 . If the sentence is indivisible and its mean¬ 
ing also is indivisible, then everything would be 

II. 76 


direct statement ( sruti ) and there would be no such 
thing as contradiction with direct statement. 

[If the sentence and its meaning are indivisible, then one 
cannot, within the sentence-meaning, distinguish between 
what is conveyed by direct statement and what is conveyed by 
juxtaposition. Everything would be conveyed by direct state¬ 
ment. There could arise no conflict and what is more, there 
would be no substitute if what is prescribed is not available.] 

It is now shown that if a sentence is indivisible, intermediary 
sentences would become meaningless. 

76. There would be no sentence-meaning (in 
the case of a big sentence) consisting of a collection 
of minor sentences requiring one another and used 
for conveying one single meaning. 

(If individual words do not exist and are meaningless, 
intermediary sentences would be in the same position. Also, 
a meaningful sentence can become a part of a bigger sentence 
and so become meaningless which is a contradiction. 

The Vrtti explains the same idea with an example : gaur 
duhyatam , upddhyayah payasa bhunktvd mdmadhydpayi$yati = 'let 
the cow be milked, the teacher will eat (his rice) with milk and 
then teach me. Here a big sentence, having an intermediary 
sentence as its part, is expressive of an action which is qualified 
by another action having its own accessory. If the parts (in 
the form of intermediary sentences) requiring one another are 
not connected with the one main meaning, then the main 
sentence would also be meaningless. Besides there is no fixity in 
limits of intermediary sentences. Sometimes, gdmabhydja= 'drive 
the cow on’is the sentence,sometimesitis: Devadatta\ gam abhyaja 
= 0 ! Devadatta ! drive the cow on. Sometimes it is : Deva - 
dattagam abhyaja §ukldm= O ! Devadatta, drive the white cow 
on! That being so, one would have to accept the contradictory 
position that the same thing is sometimes meaningful and some¬ 
times not. The first sentence is meaningful, while the same 
words are meaningless in the next, being only a part thereof.] 
It is now pointed out that, if the individual word and its 
meaning are denied any existence, certain principles followed 



in the world and in the Sastra would become inexplicable. These 
principles form the subject matter of the Mimamsd-sidras of 

Jaimini. _ 

77. That this action is prasahgika, this is ob¬ 
tained through tantra (extension), this through 
avrtti (repetition) or bheda (difference), there is 
badha (suspension) here and combination there 
( samuccaya ). 

[The principle of prasanga is established in the twefth 
adhyaya of the Mi.Su. Madhava, in his Jaiminiya-nyayamala- 
vistara XII. 1. defines it thus:— anyoddesena anyadiyasyapi sahd- 
nu$thdnam prasangah= ‘the single performance of a subsidiary 
action, accepted as helpingaprimary action other than the 
one to which it belongs.’ For instance, prayaja and anuyaja 
offerings, taught as subsidiary to the agtifyomiy a animal sacrifice, 
serve as subsidiary to the cake-offering also. See Mi. Su. XII.I, 
1-6. This principle is sometimes followed in the world also. 
When a teacher is teaching one student and other students arrive 
and profit by the same teaching instead of asking the teacher to 
instruct them separately, it is a worldly example of the principle. 
It is sometimes followed in the Vyakaranasastra also. For exam¬ 
ple, P.1.1.27 not only gives the name sarvanama to some words 
but tells us incidentally that in the word sarvanama, nis not 
changed to n, as it normally should. If individual words have 
no meaning, this principle cannot be applied. 

The principle of tantra is established in chapter XI of the 
Mi. Su. It is similar to the principle of prasanga. The differ¬ 
ence is that in tantra, the single performance of a subsidiary 
rite is prescribed and it is intended by the sacrificer to help 
more than one primary rite, whereas in prasanga no such pres¬ 
cription or intention is discernible. The prayajefti, performed 
once before or after, serves all the six sacrifices. An every-day 
instance of the application of such a principle is that of many 
students using one lamp for their study. Panini’s use of the 
word tapara in P.1.1.70 in two meanings is an instance of tantra 
in grammar. The two meanings are : (1) tab. paro yasmdt 

so'yam taparah, (2) tatparah taparah. For a Vedic application of 
this principle, see Mi. Sit. 1, 14. 

Each of these principles has its opposite which also goes 

II. 78 


to prove the existence of the individual word. Avrtti or repe¬ 
tition is the opposite of tantra— extension. A Vedic instance of 
avrtti is the statement that there are seventeen mantras for kin¬ 
dling the sacrificial fire (saptada§a samidhenyo bhavanti ). This 
number is reached by repeating the first and the last of eleven 
three times. Bheda or difference is also an extension of tantra . 
A Vedic instance of it is graham sammar^ii— he cleans the 
vessels. As there are many vessels, each one has to be cleaned 
separately ( bheda). The cleaning of one would not do for all. 

Badha or suspension is the subjectmatter of adhyaya'K ofMf. 
Sit. An instance of it would be the prohibition of the eating 
of the domesticated fowl {abhaksyo grdmyakukkulah) which would 
otherwise be possible, considering that it can also satisfy hunger. 
An instance from the world would be : After the statement: 
“Give curds to Brahmanas,” it is said c and butter-milk to 
Kaundinya 5 . The latter statement suspends the operation of 
the previous one as far as Kaundinya is concerned. The oppo¬ 
site of this principle is samuccaya. Instead of one operation 
cancelling another, both become applicable to the same case. 
An instance of it would be the statement : Feed Devadatta with 
salt, ghee and vegetables. All are given to the same person. 

These principles would not be sound if the individual word 
and its meaning did not exist.] 

78. That uha (modification) in this matter 
is justifiable, the relation here is not barred, that 
this is an instance of general transference and this 
of particular transference. 

\TJha is dealt with in chap. IX of Mi. Sft. In the Veda, 
this principle consists in the modification of the mantras pres¬ 
cribed for a prakrti-yaga in order to suit the circumstances of the 
Vikrtiyaga. In the agjieya sacrifice which is a prakrti , one has to 
say : agnaye tva ju$tam nirvapami , when one offers vrihi to Agni. 
In the Surya sacrifice which is a vikrti of it, one has to substitute: 
Suryaya tva, while offering nivara (wild rice) instead of vrihi 
(cultivated rice). Thus, we modify the mantra to suit differ¬ 
ent circumstances. The opposite of this principle would be 
not to make any modification. For instance, in the Vedic sen¬ 
tence yajamanam dandena dik$ayati , mekhalayd dik$ayati , the word 



yajamanam is joined on to each one ofthe verbs without its case- 
affix being changed. It is clear that to pick out one word of a 
mantra and substitute another for it is an admission that the 
individual word has a meaning of its own. 

Transference or atidesa is the subjectmatter of adhyaya VI I 
and VIII of the Mi. Sii. It means the transference of the details 
ofthe prakrtiyaga to a vikrti-yaga. as, for instance, those of the 
Darsapurnamasa ifti to other if (is. In everyday life also, such 
transference often takes place, as,for instance, when one says: 
Behave towards this kfattriya as towards a brahmana. In the 
Science of grammar also, there are many cases of such trans¬ 
ference. For instance, the siitra: Sthanivad adeso’nalvidhau 
(P.1.1.53.) actually teaches such transference. As such, trans¬ 
ference involves the removal of a word or its meaning from its 
original context and placing it in a new place. 11 goes to prove 
the existence of the individual word and its meaning.] 

79. That here candidature (for the fruits of 
the sacrifice) means the fact of being entitled to 
perform it, whereas no differentiation exists there 
in regard to the fruit, that one who gets his right 
to perform an action through the fastra is prohibited 
from performing another. 

[Only the sdstra can tell us who is entitled to perform 
a ceremony and who is not. This subject is discussed in the 
sixth chapter of Mi. Su. The three factors which confer the 
right are : (1) Desire for the fruit, (2) Capacity, (3) absence 
of prohibition by the Sdstra. Only the Sdstra can determine 
capacity or the absence of it in regard to invisible matters.] 

80. T hat here the sequence is understood through 
direct injunction, there through utterance etc., that 
the sequence is important here, insignificant there. 

[How to determine the order in which the different minor 
rituals belonging to a major sacrifice are to be performed is 
discussed in Adhyaya V of the Mi.Sii. Many guides are given 
there. In the sentence: hrdayasyagre’vadyati, athajihvayah, the 
two words agre— first and atha — ‘afterwards’ indicate the order. 
In the sutra : parasmaipaddnairi nalatus .. (P. 3.4.82), the order of 
enumeration indicates the order of application. 

II. 81-82 


This order is sometimes very important, as, for instance m 
the Vedic sentence : pratliamam bhojayitavyah tato ’ bhyan- 
y j anam = he should be fed first and then anointed. Sometimes, 
it is quite significant. Thus, though a tiaimittika sacrifice maybe 
mentioned after a kdmya one, the two need not be performed 
in their order, because the two are not connected together by 
one purpose. Therefore, if the occasion ( nimitta) arises earlier 
than the desire, the naimittika sacrifice may be performed earlier. 
See Mi.Su. V. 3. 32-36. All this is possible only if the individual 
word and its meaning have a separate existence, because, the 
determination of the order involves picking a word out of a 

81. That this is connected with the subsi¬ 
diary actions of another and, therefore, brings into 
existence the subsidiaries of the ritual in question, 
that this does bring into existence (the subsidiaries), 
that here, this is inevitable. 

[The motive of actions is discussed in adhyaya IV of Mi. Su. 
For the principles referred to in the present stanza, examples 
canbefound (1) in theworld, (2) in the Vedas, (3) in Vydkarana. 
Punyaraja gives only one example. The king who sits on the 
elephant causes the umbrella to be held over him and thus 
brings into existence its shade. It is not for the elephant that 
the umbrella is held and so it does not bring it into existence, 
(aprayojako), but as it also benefits by the shade, it also may be 
said to bring it into existence. Both the king and the elephant 
enjoy the same result, namely, the shade. This is an example 
from the world. An example from Vydkarana is this : P. 4.1.92 
and P. 4-1 -95 produce the same result, namely, the form dakfi= 
son oidakfa. In these two examples, two things become prayo- 
jaka, because they lead to the same result or benefit by the 
same result.] 

82. That this is primary and this is secondary, 
that this is the mode of procedure, that this is directly 
useful and this is indirectly useful. 

[Much space is given in Mimarrisd to the question of what is 
primary and*what is subsidiary. In the process of threshing 



the corn, for instance, the threshing is subsidiary to the corn 
to unhusk which it is performed. Among the different actions 
constituting a Vedic sacrifice, some are directly useful while 
others are only indirectly so. For instance, the threshing 
of corn is directly useful for the performance of the Daria 
purnamasa sacrifice while prayaja is so only indirectly. In 
Vyakarana also, this distinction can be found. For instance 
the stem is directly useful to the suffix; that which qualifies the 
stem as a preposition is only indirectly so. An example from 
the world is : the ornaments which a man himself wears are 
directly useful to him, while those worn by his wife and children 
are only indirectly so. This distinction between primary and 
secondary, directly useful and indirectly useful, presupposes the 
existence of parts of sentence and sentence-meaning.] 

83. That here there is difference of power as 
well as of function, while there, there is a difference 
in the result, that here the distinction has arisen 
out of the (particular) relation while there the 
difference is insignificant. 

\Bheda or difference among actions is dealt with in adhyaya II 
of the Mi.Su. As in the other cases, difference can be illustrat¬ 
ed also by examples taken from the world and the vyakarana- 
Sdstra. Lightning can be described as follows : valahakad vidyo- 
tate=‘ it flashes from the cloud’, valdhake vidyotate= ‘there is a 
flash in the cloud’ valdhako vidyotate=‘ the cloud flashes’. In 
these three sentences,the cloud is described as the starting point 
(apadana), the abode ( adhikarana ) and the agent ( karla ) of the 
act of flashing. There is difference in power and, according to 
that, grammar prescribes different endings. This is the differ¬ 
ence in function. Sometimes, there is only difference in power. 
For example, in dhanufa vidhyati= ‘he hits with the bow’. Here 
there is only one case ending, the third one but there are two 
powers. The bow is thought of as an instrument of the act of 
hitting but it cannot be so unless it is the starting-point for the 
departure of the arrow. The bow is both apadana and karana, 
two distinct powers, but the former is not expressed. The 
making of a gift with a view to attain long life, health and pros¬ 
perity is an instance of difference in result. When different 

II. 84-85 


suffixes expressing different agents are connected with the same 
root, the action expressed by this root also appears to be diffe¬ 
rent. For example, pacati , pacanti. In the sentence pakivaudanam 
bhunkte-= ‘he cooks the rice and then eats it’, there really ought 
to be a difference in the actions performed by the agent and the 
object, but this difference is not emphasised. Hence, there is 
only one agent for both actions.] 

In addition to the foregoing arguments, the Mimamsaka 
wants to put forward some others based chiefly on vyakarana, 
in order to prove the existence of the individual word. 

84. That here the negative particle is connect¬ 
ed with the verb while there it is connected with 
the noun, that this is secondary while that is primary, 
that this is pervasive while that is long or short. 

[ Prasajyapratifedha means a negation in which the negative 
particle is connected with the verb. Paryudasa is a negation 
in which the negative particle is connected with the noun. 
Brahmano na hanlavyah is an example of the former and the siitra : 
dto’nupasarge kali (P. 3.2.3.) is an example of the latter. In the 
siitra : tatpurus ah samdnadhikaranah karmadharayali (P. 1.2.42.) 
the word samanadhikarana qualifies tatpuru,sah. This is a sec¬ 
ondary use of the word, because, primarily, it is not the tatpuru}a 
but the component words which ar t samanadhikarana. In ekasruti 
durat sambuddhau (P. 1.2.38.), the word satnbuddhi should be taken 
in its pervasive ( vyapi ) worldly meaning. All this is possible 
only if words are taken out of the sentences and that means that 
they have a separate existence.] 

85. That this is secondary to many severally 
while there the options are numerous, that this is 
restricted while that becomes entitled to such a 
thing under these circumstances. 

[When one verb is connected with many nouns in a sen¬ 
tence, we have an instance of one thing being secondary to many 
severally. The options sometimes allowed in the application 
of the rules of Panini are explained by commentators in their 
remarks on the siitra: na veti vibhdfd (P. 1.1,44.) An example 
of restriction in Grammar is that contained in the siitra : 



patih samasa eva (P. 1.4.8.)which means that the word pati gets 
the name of ghi only when it becomes one of the members of a 
compound. When one syllable in a word gets the udatta accent, 
the others become entitled to nighata *= suppression of the acute 
accent, by the sutra: anudattam padam ekavarjam (P. 6.1.158.) 

86. That the distinction here is understood 
from indications found in another sentence while 
the meaning of the word there is obtained by ana¬ 

[When one reads of aktdh sarkarah= ‘sugar mixed with fat’ 
in the Veda, one wants toknowmixed with what fat. From the 
sentence : tejo vai ghrtam= ghee is indeed splendour, one under¬ 
stands that the sugar is to be mixed with clarified butter ( ghrta). 
This is an instance of a distinction being understood from an 
indication found in another sentence. When one says : 
rdjapuiufah, to ask ‘of which king’ and to get the answer ‘of king 
Sudraka’ is an instance of analysing the meaning of a sentence.] 

87. Such properties of sentences which depend 
upon the meaning of individual words would be 
incompatible if the individual words were considered 
to be inexpressive ( avacaka ). 

The above arguments of the Mimamsaka are now answered. 

88. What has been pointed out above is not 
inexplicable if one analyses a big sentence into 
minor sentences, even though its meaning is really 

[The idea is that just as a big sentence is divided into 
small sentences on the ground that these latter express minor 
ideas, in the same way, though the meaning of a sentence is 
really indivisible, we can recognise within it the meanings of 
individual words also. Once this is done, all the objections 
raised by the Mimamsaka are answered. 

The upholder of the individual sentence may argue in one 
of the two following ways :—>■ 

Even if meanings of individual words and of subordinate 
sentences exist, one does understand a sentence-meaning in 
addition to them. It does exist as a cognition, if not in reality. 

II. 89-91 


Such a sentence-meaning can be analysed on the basis of differ¬ 
ence in power (fakti) and the meanings of individual words 
and of subordinate sentences obtained. The Buddhists believe 
in the svalakfana only, something which is absolutely unique, 
having nothing in common with anything else in the world. 
And yet, they talk about jdti on the basis of apoha, that is, differ¬ 
ence from everything else. In the same way, the upholder of 
the sentence believes in its indivisibility. And yet, on the basis 
of analysis through difference of power, individual words and 
subordinate sentences and their meanings are recognised as 
taking place in cognition, though not' having reality.] 

An analogy is now given for seeing differentiation where 
there is really unity. 

89. Just as the same scent appears to be differ¬ 
ent when it is found (in different objects like) 
flowers, in the same way, differentiation of meaning 
can be made within the sentence also. 

[One speaks of the smell of a flower or of sandal-wood as 
though they were totally different from each other, not realising 
that, they are really the same. The idea here seems to be that 
scent is one and all-pervasive, even though it may seem to be 
different according to the substance where it is found.] 

Another analogy is now given. 

90. It is like the perception in a gavaya or in a 
man-lion, grasped by a single cognition of a part 
that is similar to a universal external to it. 

[ In a gavaya, there is no universal called gotva, in a man- 
lion (narasiniha) , there is no universal called naratva or simhatva. 
Whatever universal there is in these two objects is totally differ¬ 
ent from these two universals. But the average man thinks 
that the universals gotva, naratva and simhatva, similar to those 
existing in a cow, a man and a lion, exist in a gavaya and nara- 
simha. They do not, but mind creates the fiction. In the same 
way, the mind creates the fiction of word-meaning within the 
indivisible sentence-meaning.] 

91. When one sees the unseen and unfamiliar 
portion (in them), it is the whole thing which has 


become unintelligible to the man of feeble intelli¬ 

[When one sees in a gavaya a part that resembles the cow 
and another part that resembles some other animal, the fact is 
that one does not see correctly the object at all.] 

92. In the same way, in sentences that become 
totally different by the addition of the word pika 
(cuckoo) and other elements,, what appears to be 
similar is really non-existent. 

[Though apparently the only difference between vanad 
vrk$a aniyatam and vanat pika aniyatam is that, in the latter sen¬ 
tence, there is the word pika instead of the word vrk$a , still it is 
held that the latter sentence is totally different from the former. 
The portion that appears to be common, namely, vanad aniyatam 
is really non-existent as a sentence is indivisible. When there 
appears to be a doubt about the meaning of one word, it is really 
the meaning of the whole sentence which is in doubt.] 

93. Just as one indivisible knowledge appears 
to resemble another indivisible knowledge in one 
part and to differ in another. 

94. In the same way, even though the sentences 
are indivisible and differ from one another com¬ 
pletely one perceives difference (between them in 

[The knowledge of blue resembles the knowledge of green 
in that both are knowledge but differs from it inasmuch as the 
contents of the two are different. Or the point might be illus¬ 
trated by taking two pictures, the colour of one of which is 
green and blue while that of the other is green and yellow. 
The two pictures would resemble each other in parts and 
differ in parts but in reality they are two different partless 
wholes. Similarly, two sentences which appear to resemble 
each other in having a common word and word-meaning are 
really two different indivisible wholes.] 

II. 95-99 


95 . How is one to determine the limits of an 
individual word considering that its form changes. 
If its limits are not determined, how can its meaning 
be fixed ? 

[ In the sentence dadhy anaya, the word dadhi has changed its 
form into dadhy because of what follows.] 

If a sentence is indivisible, how can the sentence Sveto 
dhavati be the answer to two different questions at the same time? 
This is answered as follows— 

96. In the other view (anyatra= that of indi¬ 
visibility) there is a kind of coalescence of forms in 
such words as sveta h; by the law of extension (tantra), 
it is to be taken as different for these two different 

[By tantra, the same word stands for both the forms. For 
the meaning of tantra, see the notes on Ka.77.) 

97 . In one and the same word (sveta) there is 
a coalescence of different forms. Though not really 
different from each other, they are separated (because 
they convey different meanings.) 

(In Ka.96, it is said that Svetah brings two separate expres¬ 
sions to the mind : Svaitah and Svetah, each of which conveys 
its own meaning. Here we are told that Svetah directly conveys 
the two meanings.] 

It is now shown that, in grammar also, one verbal form 
sometimes conveys two things. 

98. In this Science, sometimes there is an equali¬ 
ty of number between the conveyor and the con¬ 
veyed and sometimes there is difference. 

99 . In the sutra : ukalo’jjhrasvadirghap luiah (P. 
1 . 2 . 27 ) it is on the basis of equality of number that 
mutual correspondence is established; but in the 



sutra : syaiasi Irlutoh (P. 3.1.33.) a difference in the 
number as between the conveyor and the conveyed 
is present. 

[The sutra P. 1.2.27 means : vowels having the duration 
of u, u and u are called short, long and protracted respectively. 
Thus, this sutra teaches three names to be applied to three differ¬ 
ent things. There is equality of number between the names 
and the named. P.3.1.33 means : let sya and tasi be added 
to roots when followed by Ir and tut lr stands for Irt and 
Irh. Thus there are two conveyors and three conveyed.] 

100. Ta in the sutra : yasyeti ca (P. 6.4.148.) 
denotes the two vowels ( i and a ) and stands for 
all the named (samjninah) through identity ( abheda). 
Because the bare thing which is understood cannot 
convey all of them. 

[Yasyeti ca (P. 6.4.148) means : i or a, belonging to a stem 
called bha is dropped when followed by i or a of a taddhita suffix 
or by long i. Ya in the sutra stands for f. and a. In the same 
way, the word §velah stands for sva-\-itah and Svetah. ] 

101. The form u (in P. 1.2.27) though in¬ 
divisible, is the cause of the understanding of three 
separate ‘u’s through separate sentences. This sep¬ 
arate application is just like the latter element 
[pararupa ) standing for itself and the preceding 
element (purvarupa) when it becomes the substitute 
for both. 

[ Panini often teaches that when two vowels meet, the latter 
should remain and take the place of both. For example in 
P. 6.1.94— ehi pararupam='wh.&n a preposition ending in a is 
followed by a root beginning with e or 0 , the latter vowel takes the 
place of both. This is an exception to P. 6.1.88 which re¬ 
quires vrddhi in such cases. The sound which comes in the 
place of both is called pararupa. It stands for both. Similarly, 
svetah stands for both : svd itah and Svetah.] 

All this has been said on the basis of abheda. It is now 
stated that Katyayana believes in bheda and not in abheda. 

II. 102-104 


102 . By rejecting the prolongation of a parti¬ 
cular portion (of a diphthong) in connection with 
protraction (pluti) and by rejecting the combina¬ 
tion of two vowels, he (the varttikakdra ) has adopted 

[The siilra : plutav eca idutau (P. 8.2.106.) = when e, ai, o 
or au is to be protracted, it is their first element, namely, i or u 
that is protracted. Punyaraja apparently quotes the follow¬ 
ing : ahgavivrddhir nopapadyate , na hy aico’vayaua akara, ikara, 
ukaro wz=This protraction of a part of the diphthongs in ques¬ 
tion, that is, e, ai, o, au, is not right because a, i, or u is not a part 
of c, ai, o, or au. This passage is not found in the vdrttikas or the 

While considering the sulra : samaharah svaritah (P. 1.2.31.) 
Katyayana says :— samahdro'coicet, nabhavat= if it is said that 
it is the vowels which are combined, it is not right, for there is 
no such thing. Thus Katyayana rejects the idea that here 
there is a combination of vowels.] 

Tiantra is the accepted view. That is now explained. 

103. Just as there are different forms for 
ardharca and other such words on the basis of the 
sequence (of their elements) in the same way, 
according to the view that they are one, different 
forms can be understood otherwise (that is, as one). 

[The compound word is ardharca (P. 2.4.31.) but its ana¬ 
lysis is : rco'rddham. Thus, the same thing has two forms on 
the basis of the sequence of its elements. Similarly, two differ¬ 
ent things can be combined into one on the basis of tantra as 
in ivetah ] 

104. The phonemes, in themselves unchang¬ 
ing, become different by taking on new powers 
when there is connected speech ( samlrita ). 

[The one word svetah appears to have the power to convey 
more than one meaning. When it is looked upon as the result 
of the unification of artificially analysed elements (that is. 



§va and itah) it acquires the power to convey another meaning. 
Really speaking, there is no unification of separate elements.] 

105. Objects, without giving up their real 
form are perceived differently as a result of some 
defect in our senses. The same is the case with 

[ A defect in our eyes makes us see the same thing in differ¬ 
ent forms. In the same way, on account of artificial analysis, 
the same woi'd can take on many forms.] 

106. As a result of the mode of pronouncing 
it, the same word appears to be different though 
no change has taken place in it. 

[The word Svetah, while remaining the same, appears to 
consist of iva and itah on account of a particular mode of pro¬ 

107. What is called saman is either the rk sung 
in a particular manner or it is the song itself. It is 
not a separate entity. These very transformed 
hymns (rks ) differ from one another according to 
the particular method in which they are sung. 

[It is well-known that the hymns included in the Sama- 
veda are found in the Rg-Veda also. They are collected together 
separately only for the purpose of singing them in a particular 
manner. This is what is meant by the sutra : gttifu samakliya 
{Mi. Sii. II. 1.3.6.)] 

The principle of tantra is again explained. 

108. Many forms, different from one another, 
resembling those of which only one will be retained, 
are, in this way ( upayat) coalesced into one. Be¬ 
ing uttered in a compressed form, it is considered to 
be cofrect in the sastra. 

[Tantra means compressing several forms into one. The 
compressed form stands for all of them. In the verse, the word 
ekase$inam is used because of the resemblance of the process 
adopted in tvetah to what happens in an ekaSe$a likcdevau which 

II. 109-111 


stands for devasca devasca . A real eka§e$a is the retention of one 
out of many identical forms. In Svetah, there are no identical 
forms, but there are many forms which are coalesced into one. 
There is only resemblance and not identity.] 

109. By accepting it (the word svetah ) to be 
a form common to different phrases, it should be 
used for conveying more than one meaning. Other¬ 
wise, such forms would not be correct. 

[ Svetah is an indivisible word having the meaning of‘white 5 
or it is a combination of Sva and itah and, therefore, divisible. 
It is really two words but because of the identity of the phoneme 
sequence, it is looked upon as one. That is why it can be used 
to convey two meanings. Ordinarily one word conveys one 
meaning only.] 

no. By means of compressed utterance, one 
correct form is obtained for sentences that are similar 
to one another. 

[Just as Panini has taught the retention of one of many 
identical individual words, he has not taught the retention, in 
the same manner, of one among many identical sentences. But 
padatantra^vakyatantra and vdkyaikase$a do play a part in gram¬ 
mar. Svetah is an example of padatantra : the use of adjectives 
and verbs in a sentence in such a manner that they can be taken 
in the singular or dual number. What is called tantra by the 
grammarians is $abda$le$a and artha$le$a is ekaSe$a . There is 
vakyaikaSe$a when a general statement sums up several special 
statements. For example : TaSo vidhatuh kathayanti khanditam— 
‘they declare the glory of the Creator as marred 5 . This 
general statement sums up the previous special statements 
found in the same verse. See Ainbdkarin on Vak. II. 110.] 

An example is now given. 

hi. Just as one and the same sound appears 
to be different according as it is produced by a 
flute or some other musical instrument, in the same 
way, different forms are reduced to one. 

[What is pointed out here is that just as one can become 




many through some conditioning factor, in the same way, 
through compression ( tantra ) many can become one as in 

The author now refutes the view that if the doctrine of 
indivisibility were true, intermediary sentences would become 

112. Just as the intermediary sentences (form¬ 
ing part of a big sentence or passage) and resem¬ 
bling words (forming parts of a sentence ) are recog¬ 
nised as separate, in the same way, these senten¬ 
ces may be recognised as separate (even when they 
do not form part of a sentence.) 

[The idea here seems to be that by resorting to analysis 
(apoddhara) we ascribe meanings to intermediary sentences 
which form part of a big sentence. When these intermediary 
sentences are independent, they would have a meaning all the 
more. So the objection raised in verse 76 does not hold good.] 

The author now points out a defect in the view that purpose 
{prayojana) is the meaning of a sentence. 

113. He who holds that the expressed sense 
belongs to the word and that the sentence denotes 
purpose cannot establish any connection between 
one sentence and another. 

[Punyaraja, in his commentary on verses 1-2 enumerates 
six views on the nature of the meaning of a sentence. That 
is, purpose ( prayojana ) is one of them. This is supposed to be 
common to all the views on the nature of the sentence-meaning. 
According to this, what is understood on hearing a sentence, 
that is, its abhidheya is not the sentence-meaning but the purpose 
to fulfil which the speaker utters it. But if the sentence has no 
abhidheya , an expressed meaning, there would be no connection 
between sentences because such connection is always through 
the expressed meaning.] 

But the defect can be removed according to the anvita - 
bhidhana . 

114. It is only verbs which have mutual 


II. 115-116 


quirement and a relation, based on such verbs is 
understood between sentences (even if purpose is 
taken to be the meaning of a sentence). 

[According to anvitabhidhana , the verb which expresses 
action brings the means ( sddhana) to the mind. Words expres¬ 
sive of the means do not bring action to the mind in the same 
manner. As between action and the means, the former is 
primary and the latter secondary. In this view, the verb 
brings the means to the mind and the relation between the two 
is the expressed meaning of the sentence ( abhidheya ). Thus 
sentences are not devoid of abhidheya and so there can be con¬ 
nection between them.] 

It might be said against the anvitabhidhana that, if the very. 
first word expresses the particularised meaning, the remaining 
words would be useless. This is answered as follows— 

115. Repetition or restatement tends to make 
the meaning of the word clearer. All the words 
belonging to a sentence being present the meaning 
of (the whole) sentence is present in each of them. 

[The defect mentioned is removed by pointing out that the 
other words would serve the purpose of restricting and speci¬ 
fying the other words with whose meaning the meaning of the 
first word is connected. It is also pointed out that even when 
the other words are not yet uttered, they are already present 
in the mind of the speaker and so they serve to determine the 
meaning of the first word, to give it a certain completeness. 

See verse 18.] . 

It is now stated how the indivisible sentence is divided by 


116. Even though the meaning of a sentence is 

without differentiation, divisions are brought about 
by the difference in the background of their authors. 
On this subject, there have been many views among 
ancient thinkers. • • 


vakyapadIyam of bhartrhari 

[What is pointed out here is that thinkers are in* 
fluenced, by the systems of thought to which they belong, in 
their view about the nature of the meaning of a sentence. 
They try to make it conform to their other doctrines.] 

117 . Others have declared that all words are 
the cause of a flash of understanding through prac¬ 
tice ( abliyasa ), even in the case of children and 
animals in their understanding of things as they are. 

[All words, of whatever kind, are the cause of a flash of 
understanding according to some. This is as true of those 
who know the language as of those who do not, like children 
and animals. That is why fixed words are used by men in their 
dealings with animals, so that they may have this under¬ 
standing through practice. After they get this flash of under¬ 
standing, they act in particular ways. What is meant by 
practice or lohg usage is the repeated use of the same word 
for the same purpose or in the same situation. It is a kind 
of predisposition. When the horse hears the sound of the 
whip, it understands something and acts in a particular way. 
The relation between that sound and the action of the horse 
is natural and spontaneous.] 

118 . That practice is not the result of agama 
(transmission of tradition in this life) . Some look 
upon it as convention. It is in the form of : this 
should be done after that. 

[The abhyasa mentioned in the previous vei'se is not the 
result of the transmission of tradition to the child in this life. 
It is something which the child has inherited from its previous 
life. The word agama in that verse may not mean anything more 
than cause. As we cannot see its cause in this life, the practice 
is called anagama. As it comes from previous lives, it is as good 
as eternal. Others look upon it as a result of convention, estab¬ 
lished by man or god.] 

Thus, it has been concluded that the sentence is indi¬ 
visible and that its meaning is Intuition which is also indi- 

II. 119-121 


visible. But, for practical purposes, we analyse a sentence into 
words and word-meanings. The nature of these meanings is 
now set forth. There are twelve views. The first view:— 

119 . All words have a meaning amounting to : 
‘something exists 5 . This meaning is the characteris¬ 
tic of the thing denoted by each word. In the case 
of words like go , they say, it is similar to what is 
denoted by such words as apurva , devoid and svorga. 

[The meaning of a word is in the nature of a generality, 
having no particular form. When we hear the words apurva , 
devala and svarga , we visualise no definite shape or form. This 
is what Punyaraja says. Kamalasila also says the same thing 
in his Pafijikd on Tattvasangraha 886. It is true that we visua¬ 
lise a form when we hear such words as gauh, aSvah etc., but 
that is due to the concurrence of the senses. From these three 
words, we just understand a meaning, without any shape or 

Why not include shape and form in the meaning of a 
word wherever they are understood ? 

120 . The perception of a particular form which 
takes place as a result of our seeing a word used 
constantly for a particular object does not come 
within the range of a word. That is the result of a 
special effort. 

[The special effort is the experience of the repeated use 
of a word for a particular object. It must be distinguished 
from the normal power of a word. ] 

The second view 

121 . Certain distinctive features are revealed 
by the words expressive of them while others under¬ 
stood subsequently are also considered to be this 


VAKYAPAD1YAM of bhartrhari 

[What is expressed directly by the word is the universal. 
Whatever else is understood with it is not the expressed mean¬ 
ing of the word. But some consider what is subsequently under¬ 
stood also as the meaning of the word.] 

This is now refuted. 

122 . The word expressive of the universal does 
not express the varieties of the particular which is 
necessarily understood when the word expresses the 
universal (primarily). 

[A universal must have a substratum. The particular is 
the substratum of it and so it is necessarily understood, but 
not its varieties and peculiarities. What is necessarily under¬ 
stood cannot be considered to be its meaning.] 

123 . An expressive word (like ghata = ajar) 
does not denote the shapes and forms of a jar as it 
rests only on the general idea (that is common to 
all the jars). One, however, inevitably understands 

[When we hear the word ghata, we understand only what 
is common to all jars and not all the possible forms and shapes 
of a jar.] 

An illustration of this inevitability is now given. 

124 . An action expressed by a word is never 
seen except with (the means necessary for) its accom¬ 
plishment. The understanding of the means [pra- 
yoga) happens subsequently. The same is true of 

the meaning of words. 

[The word prayoga which means execution or accomplish¬ 
ment is understood by Punyaraja in the sense of association 
with the means of an action. The idea is that when one thinks 
of an action, one necessarily thinks of the means of its accom¬ 

II. 125-126 


The third view 

125. Others accept as the expressed meaning 
of a word both the fixed actions as well as the 
means necessary for their accomplishment. 

[In this verse, the word prayoga seems to be used in a 
different meaning. It cannot mean sadhana as in the previous 
verse, because it is mentioned separately. Punyaraja does not 
discuss it. It appears to mean action or application. What is 
emphasised here is that all that is understood from the word, 
action as well as means of action, the universal as well as 
the particular, is its expressed meaning. There is no ground 
for distinguishing between what is expressed and what is 
implied. The different elements may stand in the relation of 
primary and secondary to each other.] 

The fourth view 

126 (ab). According to others, the particular 
forms taken as a whole but without choice or combina¬ 
tion are the expressed meaning of a word. 

[ If a word denotes all the forms and shapes taken on by 
the individuals coming under it, it would always have to be 
put in the plural number. If, on the other hand, it denotes 
an indefinite number of such forms and shapes, there would 
be indefiniteness in regard to number. That is why the author 
says : avikalpasamuccayah . What it appears to mean is that 
the exact number of forms and shapes is not present to the 
mind. See Tattvasahgraha , 887, with Paiijika. Punyaraja says 
that this matter will be discussed in detail in the upamdsamud - 
deSa. There is no such samuddeta among the fourteen which 
make up the third kanda . Can it be a lost samuddeia ?] 

The fifth view 

126 (cd). Still others think that it is the unreal 
connection (of things with their universal). 


VAKYAPAD1YAM of bhartrhari 

[The view is that a word like ghata denotes the relation 
between the object and the universal etc. which inhere in it. 
As this relation cannot be perceived apart from the things 
which it unites, it is said to be unreal. See Tattvasangraha 887 
with the Panjikd .] 

The sixth view 

127 (ab). The real, conditioned by the unreal, 
is the meaning of the word. 

[See Vak. III. DravyasamuddeSa where it is declared that 
all words denote the ultimate Substance as conditioned or 
limited by some unreal thing, just as all gold ornaments stand 
for gold, conditioned by the particular shape of that orna¬ 
ment. See Tattvasangraha , 888 with Panjikd. ] 

The seventh view 

127 (cd). Or the word itself, when it becomes 
the abhijalpa , is its meaning. 

[What abhijalpa is becomes clear from the next verse. 
See Tattvasangraha , 888, with the Paiijika .] 

128. When the form of the word is identified 
with the object, in the form : ‘that is this’, the word 
is called abhijalpa. 

[The word is superimposed on the object. The object 
is, as it were, hidden by the word. The two are identified. 
This identification itself is called abhijalpa and the word which 
is superimposed on the object is also called abhijalpa. See 
Tattvasangraha, 889, with the Panjikd. It is the word which 
is superimposed on the object and not vice versa. That is why 
the object is said to be hidden.] 

129 . Word and meaning, being thus invariably 
identified with each Other through long usage, one 

II. 130-132 


of them assumes predominance over the other on 

130 . In the world, a word is known chiefly as 
identical with the meaning (object): in the Science 
of Grammar, on the other hand, there can be im¬ 
portance of both, according to the speaker’s 

[Word and meaning are so intimately united that to 
separate them and to consider one of them to be more 
important than the other is difficult. In the world, the mean¬ 
ing (object) is more important than the form of the word. In 
Grammar, sometimes the word as in P. 4.2.33. and sometimes 
the object as in P. 4.1.92. assumes importance.] 

The eighth view 

131. Either because it (the object) has no 
power or because it has all powers, it is through 
words that it is presented in a fixed form such as 

[An object is as the word presents it. A word can present 
it emphasising its action aspect or its universal aspect oi its 
qualities. The object has no power of its own. Itisasthe 
word presents it. 

The ninth view 

Another way of looking at it is that a thing has all 
powers, but the word emphasises one of them according to 

The tenth view 

132 . Others declare that the meaning is some¬ 
thing mental but resting on an external object and 
is looked upon as the meaning of the word when it 
is externalised. 



[See Tattvasangraha 890 with th z Paiijika. There it is 
pointed out that as long as the meaning or the object is purely 
rtientalj it does not become connected with action. A word 
must convey something capable of being connected with action. 
So what is mentally grasped must be externalised. It is only 
then that it becomes the meaning of a word.] 

The eleventh view 

133. The external forms conveyed by some 
words are based upon distinct reminiscences. The 
meaning of others is in the nature of bare under¬ 

[Words like ghata , pata, go, convey a meaning having a 
shape which is the residual trace of the actual experience of 
the corresponding external object having a shape. Words like 
apurva, devata, and dharma convey a meaning which is not 
characterised by any shape or form. A bare understanding 
takes place.] 

The twelfth view 

134. Just as our senses perceive the same object 
in different ways, in the same way, an object is 
understood from words in different ways. 

[Punyaraja attributes the difference in our way of 
perceiving the same thing to defects in our senses. It may be 
due to other causes such as distance and the absence of suffi¬ 
cient light.] 

135. The meaning of words, intended by the 
speaker to be one thing, is understood by different 
listeners differently, according to their own back¬ 

[Punyaraja points out that everybody, in using and 
understanding a word, is influenced by his own background. 
A VaiSe$ika may use the word ghata to convey a whole but the 

II. 136-140 


Sankhya will understand from it a mere combination of qualities 
and the Jaina and the Bauddha a collection of atoms.] 

13 6. With regard to the same thing, one’s 
views undergo change. The same person sees the 
same thing differently at different times. 

[Punyaraja attributes the difference to the study, by 
the same person, of different Sastras at different times.] 

137. To one and the same word are attributed 
many meanings by one and the same person or by 
many persons according to undefined ciicum- 

[The same person understands different meanings from 
the same word at different times according to the different 
disciplines under whose influence he comes. On the other 
hand, many persons, trained under the influence of different 
systems of thought, understand different meanings from the 
same word at one and the same time.] 

13 8. Therefore, cognitions and words of people 
who have not seen the truth, being full of mistakes 
and deceptions, are unreliable. 

139. Nor can the vision of the sages, based on 
truth, be brought into human transactions, foi it is 
not the basis of the use of words. 

[In the empirical stage, the sages are on the same level as 
ordinary men. They perceive things with their senses and 
the mind and use words according to what figures in theii 
mind. See Vak III. Sam 53.]. 

140. The sky is perceived as a surface and the 
firefly as fire. There is no surface in the sky nor is 
the firefly fire. 


VAKYAPADlYAM of bhartrhari 

[ Tala is explained in the Bhdmati and in the Ratnaprabhd 
as a big frying pan made of sapphire. Naiva casti talarri vyomni = 
there is no tala in the sky leads us to take the vat in talavat as 
a matup suffix and not as the suffix vati. If this is correct, tala 
may mean nothing more than a part or division. See Ambd- 
kartri on this stanza. See Mahdbharata , Santiparvan , adh. 112 for 
this verse and Nilakantha’s commentary thereon.] 

141 . Therefore, the wise man should examine 
by reasoning even an object apprehended by direct 
perception. He should not form his idea of the 
object on the authority of the perception itself. 

[An object is not necessarily as we see it. A little re¬ 
flection may convince us that it is otherwise.] 

142 . In regard to objects which are difficult to 
define, the wise man should not deviate from the 
definitions of them adopted by men of the world in 
their usage. 

[According to Punyaraja, what is emphasised here is that 
the indefinability of worldly objects being understood, it is 
better to follow in practical life the worldly conception of 
them. To try to have another worldly conception of them 
would be useless repetition ( piftape$ana ) of work already 

The author now continues his statement on pratibha. 

143 . When the meanings (of the individual 
words) have been understood separately, a flash of 
understanding takes place which they call the mean¬ 
ing of the sentence, brought about by the meanings 
of the individual words. 

[Even though the meanings of the individual words are 
not real or rather have only a practical reality, they serve 
the purpose of bringing the sentence-meaning to the mind. 
They are the manifesters of the sentence-meaning.] 

II. 144-147 


144. It cannot be explained to others as such 
and such. It is experienced by everyone within 
himself and even the subject (of the experience) is 
not able to render an account of it to himself. 

[The difficulty of defining the special taste of a drink 
made up of many ingredients is usually given as an example 
of the difficulty of defining the meaning of a sentence.] 

145. It is something indefinable ( avicarita ) and 
it brings about a kind of amalgamation of the 
meanings of individual words, covering the whole 
sentence as it were, it becomes its object. 

[Even though this flash of understanding called sentence¬ 
meaning is indefinable, its affect can be indicated : it brings 
about a kind of amalgamation of the meanings of the indivi¬ 
dual words. One can also say that it is the meanings of the 
individual words which manifest it. Otherwise, the under¬ 
standing of the meanings of the individual words before the 
final understanding of the sentence-meaning would be useless. ] 

146. None can avoid in one’s activities that 
(flash of understanding) produced either through 
words or through the working of one’s predisposi¬ 

[All activity of living beings is preceded by this pratibha, 
which is either produced here and now or is inherited from 
previous births. It is words which we hear from others that 
produce it here and now (in addition to other factors). In 
the case of children and animals, they are born with a predis¬ 
position. Pratibha and itikartavyata are not the same. The 
latter is the result of the former.] 

147. The whole world considers that to be the 
authority (in daily life). Even the activities of 
animals develop because of that. 



[Punyaraja compares this intuition to the conscience 
of good people which is able to decide what is right and what 
is wrong quite instinctively. He quotes Kalidasa’s famous 
verse : satarn hi sandehapade$u vastusu pramanam antahkarana- 
pravrttayah = in matters of doubt, one’s own conscience is the 
guide of good people.] 

148 . Just as some substances acquire the power 
to intoxicate and the like by mere maturity, without 
the help of any special effort, in the same way are 
intuitions produced in those that possess them. 

[The Intuitions are caused, says Punyaraja by predis¬ 
positions, peculiar to every living being of every species. 
Instead of mada, another reading is manda. Punyaraja obviously 
had mada which seems to me to be a better reading.] 

149 . Who transforms the voice of the male 
cuckoo in spring ? Who teaches living beings to 
build nests etc. ? 

[It follows from the way in which the subject is treated 
that in the case of human beings also, the Intuition pro¬ 
duced at the moment when we hear words is not only caused 
by the words but also by something inherited from previous 

150 . Who goads beasts and birds on to actions 
like eating, loving, hating, swimming etc. associated 
with particular species and pedigrees ? 

[Plavana can mean floating, swimming, jumping etc. 
which animals and birds do well instinctively.] 

151 . This Intuition is the result of Tradition 
(agama) accompanied by bhavana. The Tradition 
is differentiated inasmuch as it is proximate or 

II. 152-154 


[The words agama , bhavand , asatti and viprakar$a are not 
clear. They can mean many things. The difficulty is to decide 
what Bhartrhari meant. Punyaraja understands by agama 
the word. Did he have the words of the Veda in mind ? 

The word, proximate or remote is the cause of Intuition. 
As such, it is assisted by bhavand , that is, the tendency to act 
according to the nature of the different classes of beings. This 
tendency is either inherited from previous births or it arises 
in this very life. From the karika , it appears that the relation 
of causality is bhavand^>dgama^>pratibhd. By bhavand , does 
Bhartrhari mean what he calls Sabdabhdvand in Vak. I. 114 
( 122 )?] 

152. This Intuition is of six kinds according 
ns it results from nature, adherence to one’s own 
Veda, Practice, Toga , Invisible factor, intervention of 
specially qualified persons. 

[The Vrtti and Punyaraja differ widely in their examples 
of svabhava . Punyaraja cites the behaviour of a monkey as an 
example of Intuition caused by Nature (svabhava). The 
Vrtti, on the other hand, mentions the natural tendency of 
Prakrti to evolve into mahat etc., our natural tendency to awake 
after sleep. The knowledge of Vasistha and others is given as 
an example of pratibha resulting from adherence to one’s Veda. 
Intuition resulting from practice is exemplified by the know¬ 
ledge of well-diggers as to the exact location of water in the 
ground. Yogis have Intuition of what is going on in other 
people’s minds. The power of Raksasas etc. to enter into other 
people’s bodies and to disappear suddenly is attributed to 
adr$ta (Invisible factor). Lastly, the knowledge which Sanjaya 
and others got of the progress of the widely scattered fighting 
in the Mahabharata war was due to the intervention of 
specially qualified persons like Krsnadvaipayana.] 

The author now begins the consideration of the question 
of what is primary and important and what is secondary and 
implied in the meaning of a word. 

153-154. Just as the word go (cow), though it 


VAKYAPADlYAM of bhartrhari 

might be applied to an animal adorned with arti¬ 
cles which cling to it, cannot be said to be expres¬ 
sive of these articles. 

In the same way, a word, though applied to objects 
which are characterised by shape, colour and parts, cannot 
include these characteristics (in its denotation). 

[A word denotes either the universal as Vajapyayana 
thinks or it denotes the particular as Vyadi thinks, though 
both are understood. Even the shape, colour etc. of an object 
are not included in the expressed meaning of a word, what 
to say of external objects like ornaments, temporarily asso¬ 
ciated with an object. 

The Vrtti says the same thing slightly differently. Both 
for one who considers the individual to be the meaning of the 
word and for one who considers the universal to be it, the 
other things which are different from the real meaning of the 
word are just understood, that’is, they are not part of the 
denotation of the word— nabhidheyatvena Srutibhili prakaSyante .] 

Something is now said about words whose primary 
meaning is form, shape or colour. 

155 . A word which is applied to an object as 
qualified by a shape, colour and parts, cannot be 
considered to denote only a portion thereof. 

[Words like sthula , hrasva , karbura , Sabala do express parti¬ 
cular shapes and colours directly. They are not merely im¬ 
plied. Some shape or colour is the very basis of their appli¬ 
cation. They do not denote a part of these shapes or colours. 
Hundred may include fifty as its part but the word Satam does 
not denote fifty. 

The Vrtti says the same thing but gives its own illustra¬ 
tions. Words like parimandala , dirgha , caturaSra denote things 
having these shapes and not parts of these shapes. Similarly, 
words like muffi , granthi , sandhi , kundala denote things having 
that shape and not parts of that shape. Words expressive of 
colour such as citra, kalma§a , sarahga do not denote parts of 
these colours. Words like iatam , sahasram y prastha , drona y md$a y 

II. 156-158 


samvatsara are expressive of wholes and do not denote their 

Apart from such words, ordinarily a word denotes the 

156. A word denoting water applies equally to 
a drop and to a large collection of it, irrespective of 
number, size and shape. 

[The M. Bhd says that the word ghrta = clarified butter, 
can be applied to a drop or to a whole gallon of it. (See 
M.Bha. I. p. 184, line 19. and on P. 5.1.115.)]. 

157. A word (like taila or ghrta applied to oil 
etc, particularised through improvement etc. really 
denotes a part of it, identified with the whole. It is, 
therefore, really a word expressive of a part. 

(Both improvement and pollution of substances like oil 
mean their particularisation. Still, we use the general word 
to denote the improved or polluted substance. Thus used, it 
denotes a part of it only, but a part identified with the whole. 
This is a reference to M. Bhd. I. p.12, 1.18-20. The context is 
the explanation of the varttika which says that Vydkarana stands 
for both the forms tc be explained ( lak$ya ) and the rules 
which explain them ( lak$ana ). And yet, sometimes, ,we apply 
the word to the rules only. In the world also, a word which 
means the whole is sometimes applied to the part. Pancdla is 
the name of the whole country but when we say ipiirve pancaldh , 
it is applied only to its eastern region. When we say tailam bhnk- 
tam , though the word taila stands for all the oil that has been 
medicated, here it means only the dose which has been taken. 
In other words, the word for the whole has been used for a 

15 8. A word, the use of which is connected with 
a particular meaning, ceases to be used if that mean¬ 
ing is absent. 


VAKYAPADlYAM of bhartrhari 

[This karika gives the fixed definition of the cause of the 
application of the word (prayojana) of what is conveyed by the 
word ( abhidheya ). 

It is given as verse 160 by Dr. R. Pillai but the Vrtti 
. gives it here. ] 

15 9. A word which is used after including in 
its denotation attributes which happen to be present 
does not invariably depend upon the presence of 
these attributes before it is used. 

[There may be certain things which are not the cause 
of the application of a word to something, but accompany the 
cause, are connected with it and are understood at the time 
of the use of the word, as though they are part of the meaning 
of the word. But their presence or absence does not affect the 
application of the word. As the Vrtti says : te$am sannidhyam 
asannidhyam vo, §Q,bddpr&vrttdvQ.kdrQjicirii .] 

160. Even though the word ‘co^w 5 may be used 
even in the absence of hair, hoof etc., it cannot be 
used when ‘cow-ness 5 itself is absent. 

[Here also, a distinction is made between that which 
is the real meaning of a word and causes its application to an 
object and that which may ordinarily accompany it. The latter 
is not included in the meaning of the word. A part is included 
in the whole but the word which denotes the whole does not 
necessarily express its parts. Similarly, a particular shape or 
colour may accompany the universal but it is not the meaning 
of the word as the universal is. 

The Vrtti points out that it is the presence or absence of 
the universal which determines whether a word can be used 
to denote an object. When the universal is understood from 
the word, its substratum is also understood. One does see 
the use of a word expressive of the original material for its 
modification ( vikara ) also. Similarly, a word expressive of 
the whole is used for a part also. In the words of the Vrtti —- 

II. 161-163 


JDrfta ca vikdre ca prakrtau ca prakrlisabdapravrtlih. Avaycive ca 
samudaye ca samudaya§abdapravrtlidar§anam. ] 

161. It is difficult for anybody to see all the 
parts of an object. From the few parts which are 
perceived, the whole object is inferred. 

[Here a kind of analogy is made between perceptive 
knowledge and knowledge obtained through words. It is 
well-known that the word expresses what is understood 
through the senses. It was said before that it expresses the uni¬ 
versal and not the things which may accompany it. In other 
words, it does not denote the whole object. Similarly, percep¬ 
tion also does not cover the whole object. 

The Vrtti points out that sometimes, after seeing a part 
of an object, one cannot infer the whole. A doubt may linger : 
kvacitta sandeho naiva nivartate. tadyatha dadhimatradarSane . 
The example, however, is not clear. The text may be wrong. 
Or it may mean that when we see curds only, we cannot 
decide from which milk (cow’s or buffalo’s, for instance), it 
has been made.] 

162. In the same way, it is seen that one under¬ 
stands (from the words jdtigandha , utpalagandlia 
etc.) those qualities which accompany and are 
always associated with the smell (of these two 

[The idea here is that though the qualities which accom¬ 
pany the smell are understood, they cannot be looked upon as 
the meaning of these two words. The Sarikhyas who look upon 
an object (substance) as nothing more than a combination of 
some qualities and not as an entity over and above them may 
hold that the words express the qualities. The Vrtti is pro¬ 
bably referring to them in its last sentence on this verse.] 

163. Therefore, even though qualities which 
happen to be present are understood from a word, 


VAKYAPAD1YAM of bhartrhari 

it is that which is invariably present which is inti¬ 
mately connected with the word. 

[A possible accompanying quality is not recognised as the 
expressed meaning of a word. A natural relation of such a 
meaning with the word is not acceptable:— j\'a hi tathdbhutenar- 
thatmana sabdasya sambanddho ’piyujyate, as the Vrtti puts it.] 

Now begins the consideration of the meaning of parts or 

164. Case-endings are either expressive or sug¬ 
gestive of numbers like two. Or it might be consi¬ 
dered that the whole (consisting of stem and suffix 
combined) denotes an object qualified by number 

165. Or it might be considered that words like 
‘cow’ denote objects possessing number according 
to their nature, without (actually) expressing this 

[Three views are expressed in the above two verses : (1) 
the case-endings express or suggest (illuminate, manifest) 
number, means etc., (2) the stem and the case-ending together 
denote number etc., either by itself being meaningless, (3) 
words ending in case-endings denote objects qualified by 
number, etc. without being actually expressive of them. Such 
discussions are the early forms of the discussions in later gram¬ 
matical literature as to whether the stem ( pratipadika ) denotes 
one, two, three, four or five things. See M.Bha. on P. 1.2.64. 

As usual, the Vrtti is written in rather obscure language 
and here and there the text is also doubtful. Its contents may 
be summarised somewhat as follows—-Just as worldly 
usage is done by dividing the sentence into words and word- 
meanings, in the same way, fastraic usage is carried out by 
abstracting stems and suffixes from individual words. In this 
matter, some think as follows—If the method of agreement 
and difference is adopted there is, in a word, only as much 

II. 165 


meaning as can be obtained by this method and nothing 
beyond that for the whole. Whatever additional meaning is 
understood from the whole does not come from the word itself. 
The Mimamsakas argue as follows—Stems and suffixes have 
been mostly put forward on the basis of analysis by the cul¬ 
tured and their meanings, the basis of the iastraic work of the 
Acaryas , are analysed by the practical method of agreement 
and difference {anvaya and vyatireka) .This analysis is for the pur¬ 
pose of showing that these two meanings are only a means of 
•conveying the whole. In the world, for the cognition of the mea¬ 
ning of the whole, there is no division into stem-meaning and 
suffix meaning. Therefore, according to some, the stem conveys 
its own meaning ( svartha) the individual (dravya), gender, num¬ 
ber and case. The case-endings only illuminate the cases like 
the object {karma) . According to others, the stem expresses 
only its own meaning, the individual, and gender and the case- 
endings denote number and case. The expression of number 
and case by the stem is only optional, according to possibility. 
According to some, gender is an expressed meaning while 
according to others, it is only illuminated. Illumination is of 
two kinds : (1) conveying something which has no verbal 
element of its own ( andvirbhutdvirbhavanam ), (2) eliminating 

one and retaining the other. For example, in prati^thate y 
utpucchayate and abhimandyate . Retaining one takes place in the 
case of words the usage of which is well-known or not well- 
known. For example, upaste prapacati y adhite , adhyeti . Or the 
stem and the suffix together convey a single idea as in pdcaka , 
gopayita , brdhmanadhma and jugupsate. In these words, the 
suffixes aka , dya y kha and san do not convey any separate 

Alternatively, the whole, undivided into parts, endowed 
with many powers, closely linked with one another, expresses 
a meaning having a number. Without directly expressing a 
number at the time of the use of the word, meanings deter¬ 
mined by different numbers which are transitory, are conveyed 
by words having different forms. 

The Vrtti says that the different numbers are transitory 
like carpets round the neck of a crow :— samkhydvi$e$aih kdka - 
kanthe gunakambalavadanityaili. The analogy is not clear.] 


VAKYAPAD1YAM of bhartrhari 

166. The meaning of those words whose rela¬ 
tion with meaning is eternal and whose power to 
denote their meaning has been understood by 
analysis, can also be understood by agreement and 

[Words and their meanings are analysed by the method 
of anvaya and vyatireka. That is, recurrent parts of the meaning 
are ascribed to recurrent parts of the word.] 

167. The meaning of a particular part of a 
word (stem or suffix) is to be taken as settled only 
if these (agreement and difference) can be applied 
without fail. That is not the case (for instance) in 
nut and sap. 

[The last point in this verse can be illustrated as follows 
—In bhavatam (genitive plural of bhavat) there is no nut, still 
it expresses the sense of the genitive plural just as well as 
devanam where, ml is present. Can we say, therefore, that 
nut has a meaning ? In the same way, in atti, third person 
singular form of the root ad, there is no Sap. But in pacati, 
there is Sap. Can we say that Sap has any meaning of its own ? 
All that we can say is that it just helps the other elements to 
express their meaning. ] 

168. Where there is a possibility (of the stem 
and the suffix) having each its own meaning, it is 
not right to include the meaning of one in that of 
the other. The powers of words, when in contact 
with one another, are fixed and dependent upon 
one another. 

[One does see that stem and suffixes can express a mean¬ 
ing when the other is absent. The word kim has no suffix 
and yet it denotes a meaning. I n"iyan,~iyat, there is ho stem 
and still they are expressive. In alet , there is no suffix to see 
and yet it is expressive. Though ordinarily, stem and suffix 

II. 169-170 


come together in usage it is possible, by using the method of 
anvaya and vyatireka, to isolate the meaning of each and ascribe 
it to it. It would be wrong to include the meaning of each 
into that of the other. 

The Vrtti says something like this :—If the stem and the 
suffix have their own meaning, anybody would accept that the 
latter is expressed by the former. To consider it as included 
in something else would not be right because that would 
affect the natural power of words. That would lead to all 
meaning being considered as included in something else 
according to one’s fancy. Therefore'whatever meaning is under¬ 
stood from a word in usage should be ascribed to that word. 
Even though when words are used, stems and suffixes are not 
used in isolation and so their meanings are found to be mixed 
up, still it is accepted that stems and suffixes are expressive of 
their meanings separately. Because the powers of words have 
been handed down in the Sastra as distinct on the basis of 
regulation arrived at by the method of anvaya and vyatireka. As 
the Vrtti puts it —pratiniyamena krtapravibhaga vyavasthitali Sabda- 
nam saktayah pratijnayante ]. 

Where analysis would not result in a distinct meaning 
for the different elements, it should not be resorted to. 

16 9. In the words kupa, yupa and supa, one 
does not see any recurring meaning (for the re¬ 
current word element). Therefore, it is the whole 
word which expresses a different meaning. 

[Kupa, yupa and supa have up a in common. The remain¬ 
ing element is peculiar to each word, but the meaning of the 
word is not due to its peculiar element. It belongs to the 
whole word. (See. M. Bha. I. p. 32, lines 2-7.) ] 

170. In the derivations of words, one resorts 
to many ways of explaining them. Where many 
meanings are possible, a particular one is taken as 
the basis of derivation. 

[The Vrtti argues as follows—When experts put forward 


VAKYAPADIYAM of bhartrhari 

derivations of words, either by giving them readymade ( nipa - 
tana) or by giving rules of word-formation, one sees various 
kinds of explanations of correct words on the basis of some 
characteristic which is either fundamental ( savyapara ) or 
secondary, the worldly meaning of the word being invariable. 
Among the many powers of an object, any one may be resort¬ 
ed to as the basic characteristic and used as help in deriva¬ 
tion : anekasaktiyukte'rthe yd kdcinnimitabhdvendsriyamdnd 

saktih sadhutvanvakliyane 5 ngatvam pratipadyate. For example, the 
derivation of the word tandula is given in the Unadisutra — 
Vrnlutitanitadibhya ulac tandasca (U.S.5.8). It is possible to 
derive the word tandula by adding the suffix ulac to other 
roots and substituting tanda for those roots.]. 

171. Words like vaira, Vasistha, gii isa and ckaga- 
rika have been explained by some on the basis of a 
variety of meanings. 

[Vaira c an be explained as virasya idam or as virdyd idam , 
Vasifthasya idam or Vasi$thena krtam proktam va can result in 
Vasistha. Girisa can stand for girav Sete or girirn Syati . 

In addition to the words mentioned in the karika itself, 
the Vrtti points out that the Varttika : tap parvamarudbhyam on 
P. 5.2.121 is used to explain parvata and marutta. Similarly 
the idea of giving or what has happened ( Vrtta ) is used to 
explain various forms. Thus various meanings and various 
limits of stems and suffixes are used by grammarians in order 
to explain forms without any restriction.] 

172. Just as there is no contradiction in show¬ 
ing the same path in relation to a tree, an anthill 
or a mountain, in the-same way, words like go can 
be derived through different accompanying attri¬ 

[The word go really denotes the universal gotva but it 
can be derived on the basis of any one of the many attributes 
which co-exist in the cow with its universal. These attributes 
may be actions or qualities.] 

II. 173-175 


17 3. People who observe the different condi¬ 
tions of the object (denoted by the word in question) 
explain a word like, say, kimsuka by taking hold of 
one particular condition. 

[Punyaraja points out that somebody might explain the 
word kimsuka which means the palaSa tree by reference to its 
state when it is devoid of all fruit and, therefore, of parrots. 
The tree is called kimSuka = what ! parrot! meaning some¬ 
thing where there is no parrot because there is no fruit. Ob¬ 
jects have many powers or states according to time and place. 
Man observes them and applies a word to it according to any 
one of these powers or states. The two become closely asso¬ 

The Vrtti points out that there is no contradiction in fix¬ 
ing the form of a word on the basis of one particular power 
as in the case of the word kimiuka for pal as a, arrived at on 
the basis of the absence of any parrot (hika) on it when there 
is no fruit : kimsukdndm iva kalabhede Saktau kasyaiicicchabdasya 
vyavasthayairi na virudhyate .] 

174. Some have derived the word go from girati 
(to utter) or garjati (to roar) or garni (to go) or 
gavati (to sound or) gadati (to articulate). 

[The purpose of the verse is to show the extraordinary 
variety that there can be in the derivation of the same word 
by different scholars. Here the word go is taken as the exam¬ 
ple. Though usually it is derived from gacchati, it is possible 
to derive it from any one of the other roots mentioned in 
the verse. Both the sound of the verb and its meaning have 
something to do with the derivation.] 

175. Others have declared that the word gauh 
is applied to a cow because of its form (and not 
because of any meaning). All words are devoid of 
any derivation. Others take both (the form as well 
as the meaning) as the basis. 




[If the meaning of the root is the basis of the derivation 
of a word, one can think of several alternative roots as the 
basis of derivation. But there was another view on the 
subject, namely, that the form of the word itself is the basis of 
the derivation. This view is attributed to the Aukthikyas. 
Words are devoid of any derivation based on meaning. This 
is the avyutpattipak$a = the view that words are not made up of 
smaller meaningful elements. The opposite view is that both 
form and meaning are the basis of the application of a word 
to an object. This is the Vyutpattipaksa. 

The text of the Vrtti is not clear in some places. It also 
connects the avyutpattipakja with the Aukthikyas : Apare cacarya 
aukthikyadayo gauh kasmad gaur ityeva gaur iti nirvacanam ahull ]. 

The next question is : The same word or sound is found 
in different contexts in the language. Is it the same every¬ 
where ? 

176. For the sake of simplicity, instruction in 
grammar is based on what is generic. The parti¬ 
cular forms are expressive of this common element 
as in the case of the other universals. 

[The Vrtti says something like this : In the Science of 
Grammar, whatever can be taught on the basis of what is 
generic, of what recurs in the form or .the meaning in the 
midst of changes is so taught. But the particular relating to 
the form or the meaning conveys another universal, as it were. 
The root gam, expressive of its conventional meaning, does 
not express the real meaning of the word gauh. The verbs 
sravati, syandate, plavate, patati are taught in the sense of 
motion in general ( gatisamanye ) also, but they denote naturally 
particular movements coming under motion in general. In 
the same way, the word gauh, also denotes a particular 
motion. In the word gauh the root gam does not stand for 
mere shifting of the legs. Roots being polysemic, there is 
nothing to prevent the root gam from denoting all movements 
coming under motion in general. It is actually seen that the 
root gam stands for other activities as in 'gurutalpaga. Thus 

II. 177-178 


from conventional words, not only is the generic meaning 
understood, but also the particular meaning.] 

17 7. The same stem used in another meaning 
is considered to be a different one. Though it has 
the same form in usage, it is not reckoned the same 
in the other case. 

[This verse further explains the idea contained in the 
previous one. A separate word is applied to every separate 
object (praly art hath iabdaniveSah). The root gam in the word go, 
standing for a particular kind of movement, is different from 
gam standing for motion in general. The root pac in tandulam 
pacati is different from the root pac in pacyate tandulah svayam 
eva. because there is a difference in meaning. The two roots 
look alike, but they ai'e not the same. 

The Vrtti also points out that a root must be considered to 
be different when it is used in a karmakartr formation. Even 
though there may be resemblance in form and meaning, 
there is difference also and so there is no trace of the conven¬ 
tional roots in the non-conventional ones and vice versa : 
Karmakartrvifayavat satyapi tulyavifayariipatve ’tyantabhedanna 
rudhivisayandmarudhifvarudhivisaydnatn vd rudhisti kascid anu- 
<tango vidyate.]. 

17 8. Iji and Taji, two different roots, each 
restricted to its own scope, are explained different¬ 
ly by different people. There is indeed much va¬ 
riety in the process of derivation. 

[Punyaraja points out that some grammarians teach the 
root yaj and its samprasarana (change of y into i) when 
followed by a kit suffix. We would then get istah, iflvd etc. 
When followed by trc or tumun, no such thing takes place and 
we get yafta, y as turn etc. Others say that the root is ij and 
when followed by trc, the i is changed into y. 

The Vrtti says the same thing but gives different exampl¬ 
es. It say's that in stall and santi the root is different from 
what it is in asti and it consists of s only and has its own 



special scope, that is, when followed by a nit suffix. All these 
views about words are of a practical character and only 
resemble reality. In fact, there is no such thing as a root. 
It is only a practical postulate resorted to by scholars : na hi 
dhaturupam paramdrthena kihcid vidyate , vyavaharastu-kailcit kri- 
yate. Some declare that the root in asti consists of s only and 
that, when followed by a pit suffix, it takes the augment a. 
Some teach the roota/»=‘to speak’, ‘to explain’ the five forms 
aha etc.] 

179. Thus one should do as in the case of 
balavaya and jitvarl. There is no contradiction in 
taking identity or difference as the basis. 

[ Vaidurya is explained as something coming from vidura. 
In reality, it comes from balavaya and is only polished at 
vidura. It is, therefore, assumed that balavaya and vidura are 
the same. Or, it may be that among grammarians balavaya 
is known as vidura, just as Varanasi is known as Jitvari among 
merchants. Likewise, Visravana and Ravana are the same. The 
Vrtti adds that grammarians consider vidura to be a substitute 
for the stem balavaya when the suffix syaii follows. What 
the Vrtti and Punyaraja say is based upon the following 
slokavarttika on P. 4.3.84:— 

Balavdyo vidtiram ca prakrtyantaram eva va / 
na vai tatreti ced bruyaj jitvarivad upacaret/l 

180. It is for such purposes as the fixing of the 
position of the augment at that roots and preposi¬ 
tions are regarded separately in the discipline. In 
reality, the root itself is like that (that is, joined to 
an upasarga). 

[The whole or unity is real and not differentiation. But 
grammar has to adopt differentiation in order to do its work. 
upasargas and roots together constitute a unity and convey one 
idea. But the sastra treats them as different in order to 
regulate the position of the augment at and reduplication and 
so on.] 

II. 181-182 


181. The word samgramayati is an instance 
where (the augment a and the reduplication) 
are taught (not before the pure root) but before 
the root and the preposition combined. Particular 
actions are conveyed by particular combinations (of 
root and preposition). 

[The forms asamgrdmayat and sisamgramayisati can be 
explained only if the augment at and the reduplication are 
taught before the root preceded by a preposition. See M. 
Bhd, II. P. 23, lines 7-8. 

The Vrtti remarks—Roots are taught as separate ele¬ 
ments, expressive oi particular actions. In the verbs nivasa - 
yati, asphotayati, jugupsate, mutrayati , the suffix lyap and redupli¬ 
cation would get their proper place when the preposition and 
the root are regarded as separate elements. It is the combina¬ 
tion of sam and gram which is taught as expressive of the parti¬ 
cular action called‘fighting’. It has been made clear again that 
derivation must take place from the root samgrdma with the 
preposition included in it.] 

182. The grammatical operation relating to 
the combined root and preposition is of an inner 
nature (antaranga ). It is such an action (conveyed 
by root and preposition together) which becomes 
associated with the means of its accomplishment. 

[Punyaraja quotes the following passage from the M, Bhd 
on Va 11. on P.1.3.1 and Vd, 5. 6. on P. 6.1.135. 

kdrakdnam pravrttir vitistakriyayam, anyathakrtvamanis - 
audane pravarttante , anyatha ca $u$kaudane. 

It is only after the action is fully determined that the 
means are employed for its accomplishment. In other words, 
the root is first united with the preposition. Thus united, it 
expresses a fully determined action and such an action is 
united with the means. As the Vrtti puts it : tasmdd viti$ta- 
prakrtirupavacya viti$takriyd tathabhutaiva sddhyd sati sadhanasam - 
bandham pratipadyate .] 


vakyapadJyam of bhartrhari 

Now the opposite view is explained. 

183. It is only when the meaning of roots, fit 
to be used, is fully determined that it is qualified. 
Before being connected with the means of its accom¬ 
plishment, an action does not attain its form at all. 

[The view expressed here is that it is only when the action 
denoted by a root is connected with the accessories denoted by 
other words in the sentence that it attains its full form and be¬ 
comes fit to be qualified by prepositions. 

The Vrtti puts it as follows—When an idea has attained 
its full form and is to be qualified and there are several possible 
ways of doing so, the qualification is actually done by words 
actually used and then the relation of qualifier and qualified 
is attained. As an action is to be brought about, its relation 
with the accessories takes place first. Therefore, before that, 
action is formless ( niratmika ) and cannot enter into the rela¬ 
tion of qualifier and qualified with the prepositions which 
only manifest what is already there. According to some 
scholars, a root first enters into relation with the words ex¬ 
pressive of the accessories : purvam dhatuh sadhanena yujyata 
ityeke$am matam. ] 

The other view is now further explained— 

184. Just as the quality of being a root and the 
object of an action are assumed on the basis of a 
future connection with the means of its accomplish¬ 
ment, in the same way, a similar process can be seen 
elsewhere also. 

[The reference here is to P. 3,1.7, which teaches the 
formation of the desiderative verbs. The suffix san is added 
optionally, to a root when it becomes the object (of‘to desire’) 
and when it has the same subject as the action of desiring. 
We are here asked to add san to a root which is a karma , but 
it becomes a karma only when san is added. What is going to 
happen is assumed here to have taken place. The same 

II. 185-187 


thing can be done in regard to association of a root with 

185/ Just as lac and other dyes applied to the 
seed at the time of sowing prove useful to the fruit 
by changing its colour, 

186. in the same way, a distinction introduced 
on account of the connection in our minds bet¬ 
ween a root and a preposition appears at the time 
that words develop. 

[These two verses are meant to explain the view that a 
root first enters into relation with preposition and then only 
with words expressive of the accessories. The analogy of dye¬ 
ing the seed with lac in order to bring about some difference 
of colour in the fruit is used here. Root and preposition to- 
o-ether express a particular action. That particularity cannot 
be expressed by the root alone, no matter with how many 
accessories it is connected. The particularity which exists in 
the action, expressed by the root and the preposition together, 
may be made cleaner by association with the accessories. In 
all this discussion, a distinction is made between mental 
connection and actual connection in speech. The former 
naturally always takes place first. The Vrtti also refers to 
this previous connection in the mind. ] 

The author now proceeds to speak about the other 
parts of speech, now that something has been said about the 
noun and the verb [iiama and dkliyata). 

187. Distinctions already existing in some 
(roots) but not expressed (bythem) are brought out 
when they come into contact with preposition like 
pra and para . 

[The view about prepositions mentioned here is that 
they only reveal ( dyotaka ) a meaning and are not expressive of 

The Vrtti refers to the author of the Sangraha in support of 
this view, namely, that the preposition reveals a meaning 



already present in it and is, therefore, not expressive of it. As 
the Vrtti puts it :— Sabdantaropcigraham antarena sambhavi 
sannalabdhaniyamoyo'rthas tam dyotako niyamayan vacakatdm atikrd- 
matiti Samgrahkara aha . According to the author of the San - 
graha, the function of the preposition is to specify one of the 
peculiarities or modifications which are all potentially present 
in the action denoted by the root. Because it specifies ( niya¬ 
mayan) it is called dyotaka , revealer, manifestor and so it is far 
from being expressive ( vacakatdm atikramati .] 

Are prepositions only dyotaka or are they vdcaka also ? 

188. A preposition is expressive of some distinc¬ 
tion in the action. It is also possible for it to be a 
manifestor of it. Or it is used as a help to give 
strength to the root (to express its meaning). 

[Thus, this verse mentions three views about preposi¬ 
tions—(1) that they are expressive (2) that they are manifes¬ 
tos of something which exists elsewhere ( dyotaka ) (3) that 
they are helps, that is, they and the roots together express a 

The Vrtti clearly says that three views regarding the upa- 
sargas are held by the Acaryas : Vacakatvam dyotakatvam sahd - 
bhidhayitvam ityupasargesu trividha pratipattir acarydnam. If, due 
to association with a preposition, a meaning which a root 
cannot have is understood, then it is said to be expressive of 
it ( vdcaka) . If it manifests a meaning which the root can have 
but does not convey, it is said to be a manifestor ( dyotaka ). If 
the root and the preposition together convey a particularised 
action, then it is like a svarthika suffix.] 

The reasoning which establishes that prepositions manifest 
and do not express is now given. 

189. The ideas of going etc. which are not ex¬ 
pressed by the bare roots stha etc. are manifested 
by the prepositions pra etc. as can be established by 
the two kinds of reasoning. 

II. 189-190 


[The root stha conventionally means : to stay, to stand. 
When the preposition pra is placed before it, it means c to go 5 , 
c to start 5 . So it is the preposition which expresses the idea of 
going. This is refuted by the present verse and the double 
reasoning referred to is as follows— 

(1) Pra§abda adikarmadyotakah 

p urvodi tapacyadidrftapra - 

(2) Tifthatir anekarthah dhatutvad 

These two kinds of reasoning 
manam and viSesato dr$tenanumanam 

= The word pra reveals the 
beginning of an action. 
Because it is the word pra. 
Like the word pra seen 

The root stha ispolysemic. 
Because it is a root. 

Like the root yaj etc. ad¬ 
mitted to be polysemic 
by both sides, 
called samdnyato dr$tenanu - 
are referred to by the Vrtti 

It is now stated that the root and the preposition 
together convey the meaning. Neither by itself can do it. 

190. When adhi and pari are not used some 
other action is expressed by the root. The .root, 
by itself meaningless, expresses the meaning together 
with them. 

[In adhyagacchati and paryagacchati , adhi and pari are 
meaningless because what these two words mean is conveyed 
by dgacchati by itself. Whatever additional meaning we 
attribute to adhi and pari can be understood from the context 
without the help of these two prepositions. They are used 
only for the sake of clarity. The root by itself is also mean¬ 
ingless. So the meaning is conveyed by the two together. 
See.Af. Bha on P. 1.4.93. 

The Vrtti points out, following the M. Bha that when 
P. 1.4.93 says that adhi and pari are meaningless, what is 
meant is that they do not convey a meaning different from 
that of the root. Whether they are used or not, the same 



meaning is understood. As a result of grammatical tradition, 
the root and the preposition are looked upon as separate.] 

As upasargas are joined to roots, so are svdrthika suffixes to 
stems. The latter are now considered. 

191. In the same way, some svdrthika suffixes, 
finding themselves in other combinations (and, 
therefore) connected with a meaningless element 
express, with the help of the latter, a meaning 
belonging to this element considered separately. 

[Yava means a certain food prepared from yava = barley. 
Yavaka. also means the same thing. Here the suffix ka is 
found in a combination ( yavaka ) which is different from java, 
a separate word altogether. It is, therefore, natural to ask : 
What is the use of ka \n yavaka ? The answer is that in yavaka, 
the yava portion is really meaningless, though the word yava 
has a meaning. This very meaning is expressed in yavaka by 
the yava portion plus the ka portion. Prakrtyarthanuvadihah = 
the prakrti in yavaka is yava which has no meaning. As it 
looks like the meaningful word yava , the suffix is said to ex¬ 
press the meaning of the prakrti . This is a mere grammatical 
tradition : arthavadbhih samsr$ta iti sastravyavahare vyopadtiya- 
manah , says the Vrtti. ] 

Now something is said about nipatas , the fourth part of 
speech recognised by Yaska. 

192. Some nipatas (particles) manifest a mean¬ 
ing, some are expressive of a separate meaning, 
some, like augments ( agamah) , express a meaning 
together with other elements. 

[It is stated in this verse that nipatas can be dyotaka = 
manifestos of meaning existing elsewhere, or vacaka , directly 
expressive of meaning or lastly, they can express a meaning 
in cooperation with other elements. Punyaraja does not give 
separate examples of each kind. The Vrtti quotes some Vedic 
passage in the course of the explanation but the text is not 
clear. Those nipatas are dyotaka which are not used except 

II. 193-195 


in association with other words. Such are ca, va and so on. 
Those are called expressive which can convey a meaning by 
themselves like i a$vat,yugapat. There is no restriction as to the 
position of the former in the sentence.] 

193. It makes no difference to the manifesting 
nature of particles whether they are used before or 
after the associated words from which they differ 
in meaning. 

[The Vrtti points out that even though, logically, the 
general precedes the particular in speech, one may sometimes 
put the particular before the general as in Gargyo brdhmana 
aniyalam = let Gargya, the brdhmana be brought or SimSapd 
vrksai chidyatam = let the SirriSapd tree be cut.] 

It is now stated why some nipatas are said to be dyotaka 
while others are said to be Vacaka. 

1 94. A suffix, though expressive, is not used by 
itself. Particles ca etc. though they are separate 
words, are not used by themselves. 

[What is dyotaka is sometimes not used by itself; what is 
vacaka is also not used by itself sometimes. The difference is 
that if a particle, even though a pada, is not used by itself, it 
is dyotaka. If it is a pada and expresses a meaning by itself, it 
is vacaka, lik e$aSvat,yugapat etc.] 

195. Even if they denote the things collected 
(and not the collection), there is no diversity 
(which is the basis for the use of the sixth case-end¬ 
ing). It is a thing which is not an entity (that is ex¬ 
pressed by the particles). Action is expressed by 
other kinds of words. 

[It might be said that if ca denotes samuccaya , then the 
noun used with it would take the sixth case-ending as it does 
when used with the word samuccaya . One says Vrkfasya samuc - 
cayah , plak^asya samuccayah . So it must be deemed to denote 



not samuccaya = collection but the samuccita= things collected, 
which, when expressed by ca is asattva , something which is not 
an entity. This is due to the nature of words (fabdafakti ). 
The main idea in this verse is that words denote meanings 
according to their nature. It is the nature of particles like ca 
to denote the asattva , the non-entity. Pacati and paka both 
mean cooking, but the former conveys cooking as asattva , not a 
concrete entity but a process, whereas the latter conveys cook¬ 
ing as a thing, an entity, not as a process. This difference in 
the power of words is natural and not the result of teaching. 

The Vrtti had the reading : tiilpadair abhidhlyate , instead 
of kriyanyendbhidhiyate adopted by Punyaraja. The translation 
follows the latter.] 

196. Only such words as denote qualified ob¬ 
jects are connected with attributes. Ca and other 
such particles are always dependent on others even 
if they denote the collected. 

[Words which denote objects in which the universal and 
other properties inhere have a certain independence and they 
can be qualified by adjectives with some qualifying property. 
As far as particles like ca are concerned, they denote by their 
very nature, non-concrete things and are always dependent 
upon the use of other words. They have no independence 
and so they cannot be qualified by adjectives like independent 
words. Therefore, though particles denote non-concrete 
collected things, they are not used independently and are not 
connected with adjectives. This is their nature. 

The Vrtti also emphasises what has been said above and 
adds that in the different types of dvandva compounds, it is the 
collection to which the collected are subordinate which is ex¬ 
pressed : karmasadhane’pi samucclyata iti samuccitopasarjanah samtic- 
caya evabhidhlyate. ] 

The author now says something about karmapravacaniya. 

197. Sometimes an action creates a relation and 
disappears (that is, it is not mentioned). Some- 

II. 197-198 


times, a relation is produced while the verb is 
actually heard. 

[Rajapurusa is an example of an action producing a rela¬ 
tion and not being mentioned. In mdtuh smarati, an action is 
mentioned and it brings the relation of mother and child to 
the mind. In vrkfain prati vidyotate vidyut=‘ the lightning flash¬ 
es towards the tree’, prati is a karmapravacaniya. It specifies 
that the relation between the tree and the flash of lightning 
is that of aim and what aims ( lakfyalakfanabhdva). It does not 
manifest the action because the verb vidyotate does it. Nor 
does it denote a relation in general because the second case¬ 
ending in vrksam which comes in place of the sixth, does it. 
Nor does it bring some other action to the mind, because it 
is not understood. What it does is to specify the general 
relation. All relation is brought about by a previous action. 
It always subsists between two things which were kdrakas in 
relation to some action. In raja purusam bibharti=‘ the king 
supports the man’ the king is the agent and the man the object 
in relation to the act of supporting. 

The Vrtti also speaks about two kinds of relation and 
•claims that it is based on the Sangraha. The two kinds are : (1) 
that which is understood when no verb is used and (2) that 
which is understood when a verb is used: tirobhutakriyapadah and 
sannih'takriydpadah. While explaining these two terms, it seems 
to quote a passage from the Sangraha. As this passage is in 
prose while some other quotations from the same work are in 
verse it appears that this famous work was partly in prose 
and partly in verse. Here the two kinds of relation are illus¬ 
trated by rdjhah par us ah and mdtuh smarati.] 

198. It is for the sake of preventing com¬ 
pounds that the sixth case-ending has been taught 
in some cases. The third case-ending comes after 
guna in order to show that it is the instrument (and 
not the object). 

[There are eight sutras which teach the sixth case-ending 
in special cases (P. 2.3.52ff). No compound can be made of 


VAKYAPAD1YAM of bhartrhari 

the two words connected by the sixth case-ending in the exam¬ 
ples of these rules. For example, P. 2.3.52 teaches the sixth 
case-ending in mdtuh smaranam. One cannot make the com¬ 
pound matrsmaranam , because the karmakaraka is here thought 
of as a general relation : karmani Sesatvena vivaksite. If the 
compound is made, the sixth case-ending would have to be 
elided. No other karaka is thought of as a general relation 
according to this rule. So the case-ending would be used accord¬ 
ing to the karaka , as in matur gunaih smarati . P. 2.3.52 teaches 
a restriction which is understood in two ways. Theifa $\kd 
understands it thus : — karm'nyeva Se$atvena vivak$ite jafthi, na tu 
karakantare. Bhartrhari understands it thus : karmani fafthyeva , 
na tu samasalugadi. Thus one can have a matur gunandm smara¬ 
nam and matur gunasmaranam . 

The Vrtti also discusses why the eight siltras prescribing 
the sixth case-ending with particular roots have been given 
and why the same purpose could not have been served by the 
general rule Safthise$e (P. 2.3.50). Its view is that it has 
been done to prevent compounding : pratipadam punar 
arabhyate , samasaprati$edhah katham syad iti. ] 

199. When the verb is not mentioned, a rela¬ 
tion created by it is specified in some places by the 

[In vrk$amprati vidyotate vidyut=‘ the lightning flashes at 
the tree’, there is, first of all, the action of flashing. Secondly, 
there is the action of aiming. The lightning aims the flashing 
at the tree. The tree is the aim. Between the aim and the 
action of aiming, there is the relation of laksya-lak$anabhava. 
This relation is created by the action of aiming which is not 
mentioned^ in the sentence. The second case-ending after 
Vrksa tells us that there is a relation. What it is and by what 
action it is brought about is specified by prati , the post-posi¬ 
tion ( karmapravacaniya) . Its function, therefore, is to specify 
the particular relation brought about by the action which is 
not mentioned. It does not manifest the action of aiming 
because this action creates this particular relation between the 
tree and the action of flashing and is not mentioned. Nor 

II. 200-201 


docs it manifest the action of flashing, because the verb is 
mentioned. It only specifies the relation already expressed in 
a general way by the second case-ending. 

The Vrtti contains a quotation defining the function of 
the karmap'-avacaniya. Can it be from the Saugraha ?] 

Why not say that the post-position brings to the mind the 
action which is not mentioned ? 

20 0. That which brings the action to the mind 
would be connected with karaka case-endings, just 
like vi (in viparilikhati) . vi is not a preposition of 
the root likh. 

[ In pradesam viparilikhati, the vi brings to the mind the 
action of measuring ( vimana) and when it is connected with 
in deles am, it is connected with a karaka case-ending, pradesam 
being a karmakdraka. Thus, it is not parilikhati which would 
be connected with pradeSam. Vi would become the upasarga 
of md and not of likh. Similarly, if the post-position brings 
the action to the mind, it would be connected with a karaka 
case-ending and there would be no need for the sutra P. 2.3.8. 
But if the view is that it does not bring an action to the mind, 
then the relation between vrkfa and dyotana would be a gene¬ 
ral one and the sixth case-ending would result. To prevent 
it P. 2.3.8. teaches the second case-ending.] 

201. In the word apratyajayan, it is seen that the 
preposition prati brings to the mind the action 
denoted by tisthati. Abhi by itself is used in the 
sense of direction in connection with the verb 
sunvati (that is, in abhisunvanti ). 

[In the word apratyajayan, the preposition prati suggests 
the action denoted by the verb tifthati. The sentence in 
which it occurs is, according to Punyaraja : Devd asurati aprati- 
tifthatah prati? tham alabhamanan ajayan = the gods defeated the 
asuras who were not firmly established. Aprati is understood 
as apratitifthatah, that is, prati brings to the mind the action of 
prati?tha. It is not connected with the root ji in ajayan. In 



abhi$unvanti, abhi by itself denotes direction. It is not that the 
root denotes its action qualified by direction and that abhi 
merely manifests it ( dyotaka ). 

The Vrtti, while explaining apratyajayan, quotes Taitti- 
riya Samhita. V, to which probably the verse refers.] 

202. When they are connected with an action, 
the name karmapravacardya is given to su, ati etc. 
which are really of a different nature in order to 
prevent the change of x into s. 

[£« becomes karmapravacaniya by P.1.4.94 and ati by 
P.1.4.95. But for P.1.4.54, su would be upasarga and that would 
have the effect of causing the change of a following s into f 
by P.8.3.65, as in sufiktam bhavata. In this sentence su has not 
the characteristic of a karmapravacaniya, because it does not 
specify a relation, but modifies an action. And yet the name 
is given to it in order to explain why a following j does not 
change into s- So we get su siktam and not su fiktam. 

The Vrtti points out that su and ati do not have the charac¬ 
teristic ( pravrttinimitta ) which would justify the name karma¬ 
pravacaniya being applied to them. And yet, it has been 
applied to them in P.1.4.94, 95. on the basis’of their form 
(svarupa) only in order to prevent the names upasarga and gati 
being given to them. The name upasarga would result^in 
loss of udatta accent and the change of s into $ : tena gatyupa- 
sargasamjiiasrayanighdladikdryam na pravartate, concludes the 

203. Once the relation of cause and effect has 
been specified by anu, the third case-ending would 
result but that is prevented by special teaching. 

[The special teaching referred to is P.1.4. 84. Hetu (cause) 
is of two kinds : it is either that which brings about some¬ 
thing or that which makes known something. By P.1.4.90, 
anu gets the name karmapravacaniya when the idea of laksana 
is to be expressed. Laksana stands for both kinds of cause. 
In vrkfam anu vidyotate vidyut, the second kind of cause is ex- 

II. 203-204 


pressed. The tree is what makes the lightning known. In such 
a case, anu gets the name karmapravacaniya by P. 1.4.90 and 
takes the second case-ending by P. 2.3.8. But laksana can mean 
the first kind of cause also and that takes the third case-ending 
by P. 2.3.23. But as there is a separate *w/ra(thatis, P.1.4.84.) 
giving the name karmapravacaniya to am when the first kind of 
cause is to be expressed, the third case-ending is set aside 
and the second case-ending comes according to P. 2.3.8. 

The Vrtti says the same thing as follows—The word am 
has been seen elsewhere to suggest the action of hearing (ntia- 
mayatikriya ) and so here it brings to the mind the action of 
hearing as the cause of the relation produced by it. The 
sixth case-ending expressive of cause should therefore be used. 
But the third has been taught as expressive of cause (P.2.3.23.) 
That would then come in place of the sixth. But then anu 
has been given the name of karmapravacaniya which takes the 
second case ending which, therefore, sets aside the third case¬ 
ending. ] 

204. It does not manifest an action, nor does it 
directly express a relation nor does it supply a verb 
but it specifies a relation. 

[This stanza states the position in regard to the karma - 
pravacaniya . In vrk$am anu vidyotate vidyut , the karmapravaca¬ 
niya anu does not manifest an action, in this case, the action 
of aiming. Such an action does not come to the mind from 
the sentence. Nor does it directly express relation because 
the second case-ending does it. Nor does it bring some other 
action to the mind as vi does in pradetam viparilikhati . All that 
it does is to specify the relation, namely, lak$yalak$anabhava . 

The Vrtti explains this verse on the basis of the sentence : 
Sdkalyasya samhitam anu pravar$at. Here the word anu does not 
bring the action of ‘hearing 5 to the mind. Nor does it imply 
an action as vi does in viparilikhati . Nor does it express an 
action because the case-ending in samhitam would become a 
Karakavibhakti which it is not. Nor does it express a relation 
created by the action of hearing. So it specifies the relation 
created by the action of hearing. It specifies that the relation 
is that of cause and effect. As it performs this special function. 



it is regarded as a separate part of speech by some. The 
concluding words of the Vrtti are worth quoting —Tasmdd 
anyaprakarasambhavad ayam ntiamayati kriyopajanitam sambandham 
avacchiiiatti. Hetuhetumatsambandho' yam , nedam sambandhantaram 
iti. Etasmacca vrttibhedat pancamam padajatam karmapravacaniyah 
kaikid deary aili pratij nay ante.] 

The author again speaks about the indivisibility of the 
sentence and of the sentence-meaning. 

205. A collection of meaningless phonemes is 
cither with meaning or without meaning. It is the 
individual word which is endowed with meaning. 
There are no parts in a word. 

[The doctrine of indivisibility is challenged as follows— 
If the individual word is a collection of phonemes, if the 
phonemes have a meaning, if the word has a meaning, if the 
sentence is a collection of words and if the collection has a 
meaning, how can indivisibility be maintained ? The varttika — 
Samghatarthavattvat(V a. 12, M.Bha.I.p. 30, 1. 24.) seems to imply 
this objection. 

It is answered thus—The phonemes are never felt to be 
meaningful. Nobody has the feeling that the meaning of the 
word is made up of the meanings of phonemes, just as one 
seems to recognise in the meaning of the sentence the mean¬ 
ings of individual words. If phonemes have no meaning, they 
cannot be parts of words because division of the sound part of 
a word must correspond to the division of the meaning part. 

The Vrtti which is none too clear contains a reference to the 
Sangraha which is said to speak about ten kinds of meaningful¬ 
ness: tad ubhayarri parigrhya dasadharthavatta svabhavabheditd iti 
Sangrahe. . . .] 

206. A collection of meaningful woids, on the 
other hand, is different, being expressive of a 
meaning different from that of the component 
words (when there is connection between them) 

II. 207-208 91 

and because of the absence of a connection (bet¬ 
ween these words.) 

[The Vrtti points out that a collection of meaningful 
words may be of two kinds: (1) one in which the meanings of 
the words are interconnected, so that the collection, as a whole, 
has a meaning different from that of the words and (2) one 
in which the meanings of the words have no connection at all 
with one another.] 

20 7. Some say that when a word is analysed 
( bhede ) its two elements, one meaningful and the 

other meaningless, do not enter into any relation with 

each other while others declare that they do, as 
there is the example of the word kutira. 

[The word kutira is analysed into two parts: kutx and ra . 
The former has a meaning while the latter has not. So they 
cannot be connected, according to some, while others think 
that ra suggests the idea of smallness and so the two can be 
joined. ] 

208. Taking compounds and words ending in 
svarthika suffixes respectively as their basis, some say 
that from elements having a meaning, a collection 
having a meaning is produced while others hold 
that such a collection is not produced. 

[The word rajapuru$a is an instance of a compound having 
a meaning and made up of elements having a meaning. From 
a word made up of a meaningful stem and a svarthika suffix, 
no collection having a different meaning is produced. 

In the Vrtti on this verse, there is a quotation from the 
Sahgraha in which words as collection of smaller units are said 
to be of three kinds: (1) Sabdanvayinah — those in which the 
sounds of the smaller units can be traced but not their mean¬ 
ings, like gaurakhara and aSvakarna , (2) arthdnvayinah= those in 
which the meanings of the smaller units can be traced, but 



not their sounds, like Srotriya and vaidurya and (3) Sabdarthanva- 
yinah.= those in which both the sounds and the meanings of the 
smaller units can be traced, like rajapurufa and nilotpala. The 
verse gives the views of different thinkers. One of them is 
called Dhyanakara, the author of the Dhydna (graha). See 
Bhartrhari, p.29.] 

209. Some collections are made up of separate¬ 
ly meaningful parts. When divided, each ele¬ 
ment is separately recognisable and meaningful. 
Some (on the other hand) have to have their deno¬ 
tative power inferred by means of agreement and 

[ Samivrksa and Dadimivrkfa are examples of the first 
kind. Samjnu and Prajnu are examples of the second kind. 
Samjnu is explained as sarhgate januni asya— one whose 
knees are joined together. The word januni , though seen in the 
analysis, is not easily recognisable in the compound word. ] 

210. It is only a meaning current in the 
Science of Grammar which is shown to prove that 
phonemes are meaningful. Pure roots, stems etc. 
have no recognised meaning in the world. 

[The varttika: Arthavanto varna dhatupratipadikapratyayani- 
patanam ekavarndnam arlhadars, ndt ( Vd. 9. M.Bhd. I. p.30,1.2.) is 
considered here. The question whether single phonemes 
have a meaning or not is considered in the M.Bhd on the 
pralyahara sfitra: ha-ya-va-rat. The conclusion reached there 
is that they have a meaning only when they happen to be roots, 
stems, suffixes or particles. Even this meaning is obtained 
by the analysis practised in the Sastra. It is not seen in the 

211. The meaning of the bare krt and taddhita 
suffixes is likewise not known in the world. Similarly, 

II. 212-214 


before the inflexional suffixes are added, a word 
ending in a krt or taddhita suffix has no meaning. 

[The Vrtti quotes va. 7 on P.1.2.45, to show that bare 
suffixes have no meaning as against the alleged opposite 
view implied in the sutra P. 1.1.19— iditdc.ii ca saptamyarthe.] 

212. The meaning which is observed as being- 
expressed by words ending in such suffixes is really 
of the same nature (that is, just assumed in grammar) 
because the context is that of meaningful elements. 

[This has been said in answer to the objection that if 
words ending in krt and taddhita suffixes have no meaning, how 
is it that such words are called pratipadika, considering that to 
be a pratipadika, something must be meaningful. The answer 
amounts to saying that the meaning ascribed to words ending 
in krt and taddhita is only the result of analysis in grammar 
and not something seen in the world, because such woi'ds, 
without a case-ending, are not used in the world.] 

213. If the word and the sentence are not diffe¬ 
rent from the phonemes, the latter would ultimate¬ 
ly become expressive by virtue of the power of 
mutual requirement. 

[The view which is criticized here is that there is no such 
thing as the word or the sentence apart from the phonemes. If 
the phonemes are expressive and if the word and the sentence 
do not exist as distinct from the phonemes, the latter cannot 
be said to be expressive and that would go against the accepted 
view that it is they which are expressive.] 

214. If a collection of phonemes, with some 
missing phoneme but still expressive of the meaning 
(of the whole collection) is not considered to be a 
different word, then it is the complete word which 
is brought to the mind by it. 



[ If i$kartaram and niskartaram mean the same thing and if 
the former is not taken as entirely different from the latter, then 
it is the complete word which is first brought to the mind by 
the incomplete word and from the complete word so conveyed, 
the meaning is understood. 

The above note is based on Punyaraja. The Vrtti is not 
clear, but it also speaks about remembering the complete word 
through the incomplete word in connection with the Vedas :— 
Chandasesvapi lopesu prasiddhivikalani yani sabdantardni tair anyvna 
avaikalyayuktahprasiddhah samudayah smaryante. It also quotes 
words from the Taittiriya Samhita I, 8.10.1. and IV. 1.10.2.] 

215. Such an incomplete word, due to a 
particular reason (here the grammatical tradition) 
reminds one of the really expressive word (that is, 
the complete word) and conveys, as though directly, 
the meaning which is a step removed from it by the 
real word. 

[It is one step removed from the meaning because it first 
reminds one of the complete word and then expresses its mean¬ 

Whether the incomplete word is looked upon as the same 
or as different from the complete word, it is clear that the parts 
have no meaning. Even if the incomplete word is looked upon 
as a different word, as long as it can convey the meaning only 
after bringing the complete word to the mind, the conclusion 
is that the incomplete word which is a part has no meaning. 

The Vrtti points out that through habit and practice lis¬ 
teners think that they understand the meaning from the com- 
lete word:— tatra paricayat pratip attar ah sakfadivavikalavayavasabda- 
janitam abhimanyante . It also compares the understanding of 
the meaning from the incomplete word to the understanding of 
the meaning from gestures and signs like winking ( akf inikoca ) 
which are also supposed first to bring the corresponding word 
to the mind. ] 

That the parts cannot convey the meaning of the whole is 
now further explained by means of examples. 

II. 216-219 


216. Just as in compound words \\kz gaurakhara 
no separate meaning exists for each term and even 
if any be understood, it is not understood from the 

217. in the same way, it is of no use to detect 
meanings for individual words in the apparently 
connected meaning that is conveyed by the different 
words comprising a sentence. 

[From compound words like gaurakhara one understands 
an animal having a particular universal. No meaning is under¬ 
stood from each term and even if it is held to be understood, 
it is not taken note of at the time of the understanding of the 
meaning of the whole. Similarly, when the meaning of the sente¬ 
nce is understood, the meaning of the individual words does 
not figure in it. The sentence-meaning is like the flavour of a 
cool drink. It is something different from the flavour of the 

The Vrtti also says the same thing and discusses the same 
words : gaurakhara and aSvakarna . ] 

218. If the parts and the whole have different 
meanings, then, in compounds such opposite attri¬ 
butes as differentiation and unity would result. 

[This verse points out the difficulty that would arise if 
the meaning of each term in a compound word is considered to 
be real. From gaurakhara as a whole, an object having a parti- 
- cular universal is understood. From each one of its terms, an¬ 
other universal would be understood and these two would be 
different from each other. It would mean that from one word 
both difference and unity would be understood which is un¬ 
sound. ] 

219. Who would think of adhi etc. as expressive 
of the means (to the accomplishment of the 
action) ? In a bahuvnhi compound, how could a 


VAKYAPAD1YAM of bhartrhari 

meaning belonging to no word actually used (that 
is, anyapadartha ) be expressed ? 

[If the divisions are looked upon as real and not ficti¬ 
tious, they could become expressive instead of the whole. In 
adhibhuvi or adhistri the adhi would express location instead of 
the compound as a whole as it is generally believed. Secondly, 
if the parts are expressive, what would express the anyapadartha 
in a bahuvrihi?. The parts cannot do it and yet that is the real 
meaning of a bahuvrihi according to P. 2.2.24. 

The Vrtti mentions the compounds antastiram and adhistri 
for discussion. One can infer that in adhistri , adhi merely con¬ 
veys the power of location ( adharaSakti ) which is favourable to 
the action in question. In a bahuvrihi compound, the whole 
conveys the meaning of a word the form of which is different 
from that of its parts :— bahuvrihau ca rupanvayayuktam avayava - 
padam antarena padantaram tadabhidheye vartamanah samudayo 

220. In the words prajnu and samjnu 5 one 
does not understand a meaning from the different 
parts. Therefore, it is the whole which conveys a 
definite meaning. 

[Thus, four arguments have been given against the ex¬ 
pressivity of the parts : (1 ) Both differentiation and unity 
cannot be understood as real from the same word, (2) adhi 
etc. by themselves cannot express the means ( sadhana ), (3) in a 
bahuvrihi , the anyapadartha cannot be understood from the parts 
(4) in prajnu and samjnu , jnu has no meaning when it is by 

The Vrtti also discusses prajnu and samjnu which convey 
a meaning only as wholes. It is only in the Sastra that these 
words are analysed and a meaning ascribed to the parts. But 
that is fictitious, not known in the world. It concludes as 
follows— tasmad anarthaka avayavah sarvatra sanghata evdrthavdn .] 

Some objections against indivisibility are anticipated and 
answered. The first objection is that if individual words and 
their meanings have no reality, then a dvandva compound would 

II. 220-223 


only denote an integrated object and cannot, therefore, take 
the plural number. It is answered as follows— 

221. Just as the word gargah is used for many 
belonging to the family of garga even though only 
one word garga is used there, in the same way, the 
whole called dvandva is expressive of many. 

[From the mere fact that a dvandva compound denotes 
many things and takes the plural number, one cannot conclude 
that the individual words of a compound and their meanings 
are real. 

The Vrtti says that the word gargah conveys a meaning 
which is grasped by a single cognition and in which parts 
thought of as one figure. Similarly, a dvandva compound seems 
to have parts similar to other independent words and conveys 
many objects grasped by one cognition.] 

The second objection is that an action would have to be 
performed at the same time to all the objects denoted by a 
dvandva compound and that is impossible. The answer is_ 

222. Just as the action of feeding is applied to 
each part, in the same way, an action is applied 
separately to the things expressed by a dvandva 

[In the sentence : brahmana bhojyantam = ‘let the brahmins 
be fed’, the action of feeding is understood in regard to the 
brahmins as a whole. But when the action is implemented each 
brahmin is mentally separated from the whole and fed separate¬ 
ly. That is exactly what happens to an action enjoined in re¬ 
gard to the objects denoted by a dvandva compound.] 

The reference by a pronoun, a part of a compound, to the 
meaning of the other part of the same compound is now ex¬ 

223. When the meaning of one of the words in 
a dvandva compound is referred to by the word tad y 
there also, the word only resembles the pronoun. 


VAKYAPAD1YAM of bhartrhari 

[In the sutra : janapadatadavadhyosca (P.4.2.124.) the word 
tad refers to the word janapada. The whole sutra is in the form 
of a dvandva compound. If janapada has no separate meaning, 
how can the pronoun tad refer to it ? The answer is that the 
word tad only resembles the. pronoun tad . Really speaking, 
there is no pronoun here at all. The word is a unity and so 
is its meaning. There is no division in it. What looks like 
the pronoun tad does not really exist at all: tasmad vyavrttabheda 
evayam vitiate*rthe sarvanamavisesitavayavasarupah Sabdo vartate> 
says the Vrtti .] 

From the sentence : dhavakhadirapalaSas chidyantam , one 
understands just one meaning which cannot be divided 
and in which one cannot trace any sequence. If that is so, 
the cutting of all the trees would have to take place at the 
same time, which is impossible. This objection is now 

224. Just as, in regard to the cutting of a 
khadira tree, the action can proceed only gradually, 
part by part, in the same way, there is an order in 
the different parts of the meaning of a dvandva 

[The idea is that order or sequence is necessary for 
worldly purposes but the words convey the indivisible sequence¬ 
less idea. 

According to the Vrtti 9 this verse is meant to answer the 
objection that if the individual word and its meaning did not 
exist one cannot explain how we do understand the meanings 
of the different terms of a dvandva compound in a certain 
sequence. The answer consists in comparing the process to that 
of carrying out an order to cut just one tree, say, a khadira . 
The cutting can be done only in a certain order, first tfiebark, 
then the inner trunk and so on. In the same way, the action 
enjoined in regard to the objects conveyed by a dvandva com¬ 
pound is carried out in a certain order though the compound 
does not mention any order.] 

If only one among the above-mentioned trees is cut, we 
still consider that the order has been carried out. How to 

II. 224-227 


explain this if the meaning understood from a word has no 
parts in it ? The answer is:— 

225. Just as actions (described by a sentence) 
relating to the whole are applied part by part, such 
is the case with the elements comprising a dvandva 

[Even though the action mentioned in a sentence may 
relate to all the elements in a dvandva compound together, yet 
it is applied part by part as is done in the case of the dif¬ 
ferent elements forming part of an ekaJesa word. 

The Vrtti says the same thing and concludes:— tastnad 
ekadesasambandhinyo'pi kriydh samudayam evdnupatanti. Avayava- 
dvarikaiva hi samudayanam kriyapratipattih.] 

226. While explaining the compound to the 
ignorant with the help of the analytic sentence 
(vigrahavakya) the teacher (that is, Panini) has 
spoken about the relative importance of the mean¬ 
ings of the different terms (obtained by analysis). 

[This verse is an answer to the question: if the meaning 
of the individual words has no reality, how does one speak 
about the relative importance of the first or second term in a 
compound. First and second terms of a compound are ob¬ 
tained by artificial analysis, done for the sake of teaching 
derivation to the ignorant. They really do not exist. 

In explaining this verse, the Vrtti quotes M.Bha. I, p. 
404, 1. 3-4.] 

22 7. Even though the meaning is a unified one, 
many views regarding the importance (of the 
different terms of a negative compound) are set 
forth in the Bhasya. Their limitations are due to 
the exigencies of grammatical derivation ( prahriya ). 

[• A-brahmana is an example of a negative compound. 



The meaning of such compounds is discussed in the M.Bha, I, 


The Vrtti points out that in the M.Bha, a preference is 
shown for the uttarapadarthaprddhdnya as far as the negative 
compound is concerned: tatha hi nahsamasa etasminneva sarva- 
parikalpandsambhavat sarvapakfopanyasam krtva kciScid eva pakfah 
parigrhitah. Tatha hyuktam—idain khalvapi bhiiya uttarapadartha- 
pradhanye sati samgrhitam bhavati. kim ? anekam iti. M.Bha. I, p. 
412. lines 1-2.] 

228. By declaring that, according to the jahat- 
svartha view, words entering into a compound give 
up all their meaning, the Bhasyakara has shown 
that in a bahuvrihi compound all the terms give up 
their meaning. 

[This verse is an answer to the objection that if indi¬ 
vidual words have no meaning, how can one speak about the 
idea of the different terms giving up or not giving up their 
meaning when they enter into a compound word (jahatsvartha 
vrttih and ajahatsvartha vrttih ). In the M.Bha on P.2.1.1-, 
Patanjali goes into the question whether a compound word 
has a meaning different from that of the words entering into 
it. The two main views on this question are represented by 
the two expressions given above. If, in a bahuvrihi, the whole 
denotes a meaning totally different from that of the parts, it is 
a further proof that the meaning of the parts has no reality.] 

229. In Grammar, sometimes, the meaning of 
a stem is expressed by a suffix when the former is 
absent while the meaning of a suffix is expressed by 
the root when the former is absent. 

[The author wants to show that individual words and 
their meanings are a fiction and exist only for the sake of 
grammatical derivation. Iyan is a word where the suffix alone 
is seen and it expresses the meaning of the stem. The word 
means ‘so much’. As taught by Panini, only the suffix is left 
here, as the stem has been elided. Ahan = ‘he killed’ is an 

II. 229-233 


instance where the suffix has disappeared, its meaning being 
expressed by the root. 

The Vrtti says the same thing with the help of other 
examples, some of which are taken from the Vcd(i.~\ 

230. The meaning expressed by two suffixes as 
in pac-a-nti [sap and jhi) is sometimes expressed by 
only one as in ad-ti. Sometimes, when both are 
absent, the root expresses it. 

[According to Panirii, in conjugation, something may 
come between the root and the suffix but not always. He has 
divided all the roots of language into ten classes on this basis. 
Whether something comes between or not, the meaning is the 
same which shows that its meaning is unreal.] 

231. Those very meanings of suffixes which are 
taught as their basis in some school of grammar 
are taken as belonging to the stem (or root) in 
some other school. 

[The Vrtti says that some grammarians looked upon all 
suffixes as svdrthika, that is, they do no more than manifest the 
meaning of the prakrti, root or stem:— tathd hi kef am cit smartr- 
tidm sarva eva pratyayah svartliikah prakrtyarthamvddina itydkhyd- 
yante. ] 

232. Being well-known, only shortened forms 
like udvami and kari are used in the sastra. All 
grammatical derivations are meant to serve practi¬ 
cal purposes. 

[Grammar has its own conventions which do not conform 
to worldly usage. In grammar, one would sometimes say 
udvami for udvamati and kari for karoti or karomi, because that 
would serve some practical purpose. Such forms are not used 
in the world. ] 

233. It is only Nescience which is described 


VAKYAPAD1YAM of bhartrhari 

in the Science of Grammar through the different 
modes of derivation. But knowledge arises sponta¬ 
neously, free from the alternatives of tradition. 

[The Vrtti here is not at all clear and that is a great pity 
because one would very much like to know what, according to 
the Vrtti, Bhartrhari means by Avidya and Vidya here. Accord¬ 
ing to Punyaraja, what is stated here is that avidya. is the 
means of attaining Vidya. Sastra is all avidya, but through it, 
one ultimately attains Vidya.] 

But bow can vidya come through avidya ? 

234. Just as the effect is not related to the 
cause in a definite manner and, is therefore, indescrib¬ 
able, in the same way, knowledge, even though 
unconnected with any special means {anakliyeya) is 
still thought of as coming from the sastra. 

[The fact is that knowledge does not really come from 
avidya , that is, from the sastra. Looking upon vidya as an effect, 
it is like all other effects. No effect is related to its cause 
in a definite form. But it comes out of it in a definite manner 
and so seems to be wonderful. Adbhutena rupena upajayate , says 
Punyaraja. In the same way, the understanding of the word 
and the sentence as unity is vidya and their division into stem 
and suffix is avidya. Their understanding as a unity takes 
place when avidya in the form of division disappears. It dis¬ 
appears really because of the rise of awakening, but as awaken¬ 
ing takes place after the study of Sastra which stands for 
avidya , one thinks that vidya comes from the study of Sastra. 

The Vrtti also describes this rise of vidya from avidya as 
something wonderful: Tathavidya . . . . kuto 5 pyadbhutaya vrttyd 
pradur bhavati. Punyaraja’s adbhutena rupena is an echo of the 
Vrttis—adbhutaya vrttyd. In fact, the expression adbhuta Vrttih 
occurs already in the karikas. Once in III. Sambandha—81 
and again in III. Ka. 17. In all these occurrences, there is a 
similarity of context. The expression is used in connection 
with the appearance of the effect from the cause, with the 

II. 234-238 


appearance of many from the one and with the realisation of 
unity from plurality. Here, emphasising the wonderful 
character of this process of vidya arising out of avidya, the 
Vrtti concludes as follows —tasmad anyatrabhyasah kriyate , 
nantariyakatayanyad eva pradur bhavatiti = one studies one thing, 
namely, the Sdstra, that is, avidya and what results from it is 
the opposite of it, namely, vidya. That is why it is a wonder¬ 
ful process.] 

2 3 5. The word conveys a meaning according to 
long grammatical usage. This long wrong usage 
appears to be natural. 

[The Vrtti here points out that the wrong usage of words 
only reflects our wrong understanding of the world around us 
which is unreal, the only reality being the ultimate one.] 

2 36. The ignorant person sees parts in the 
primordial atom. Likewise, he sees its parts also as 
wholes endowed with parts. 

[The Vrtti points out that, due to our experience of all 
objects as having parts, some think of the atom also as having 
parts and indulge in discussions as to whether an atom is 
connected with objects through one of its parts or through all 
its parts : paramdnur ekaddena va sambadhyete, sarvdtmana veti .] 

23 7. By r our seeing jar and other objects, the 
universe also seems circumscribed. Due to the fact 
that objects are created, even the eternal Brahman 
appears to have had a beginning. 

238. These sastras which are a means (of attain¬ 
ing knowledge) are really misleading to ignorant 
people. But by following the unreal path, one 
attains truth in the end. 

[The Vrtti reminds one that the unreal is the means of 
attaining the real. Similarly the meaning obtained by analysis 



(apoddhara) is the means of understanding the integrated 
meaning ‘.—Vindpoddharena ndsti sthitalakfanasydrthasya firatipattir 

The author again denies the reality of the individual 
word and its meaning. 

239. As the words are (gradually) grasped, one 

undeistands the meaning in one particular way but 
when the whole sentence is grasped the same 
meaning appears to be quite different. 

[The Vrtti points out that this verse only states a general 
idea and that its illustrations are given in the following verses 
The general idea is that one sets aside at the sentence stage a 
meaning understood at the stage of the individual word If 
the latter were real, that could not happen, the relation bet 
ween the word and the meaning being eternal : nityatvdcca 
iabdarthasambandhasyayadupattam sabdena tasya punaraiakyah parit- 
yagah kartum iti.'] 

240. Those sentences in which, after manv 
meanings (are conveyed by the individual words) a 
negation follows, (in these sentences) these mean¬ 
ings are discarded and should not be taken as real. 

[In the sentence dhava-khadira-palasdS chedamyd na 
the meaning understood until the last word is heard is that 
the trees dhava, kliadira and paldsa should be cut. But as soon 
as the last word is heard just the opposite meaning is under¬ 
stood which shows that the meaning of the individual word is 
unreal. It is better not to take it seriously even before the 
last word is heard.] 

241. The sentence : ‘tree, there is not’ is the 
cause of our understanding a particular kind of 
absence. There is no connection in our mind bet¬ 
ween the negative particle and the object. 

II. 241-243 


[ If each word has its own meaning, the first word c tree 5 
(in the sentence ‘tree, there is not’ = vrkfo nasti ) conveys the 
tree as something which exists and the negation which 
follows denies its existence. If the tree exists, its existence 
cannot be denied. If it does not exist, its negation is un¬ 
necessary. In either case, the word expressing negation is 
useless. If the sentence is taken as a whole, this difficulty 
does not arise.] 

242. If it is maintained that the idea of the 
existence (of the tree) takes place in isolated under¬ 
standing ( vicchedapratipattau ) how can an idea not 
conveyed by a word disappear (simply because of 
its connection with negation) ? 

[ The idea which is refuted here seems to be this :— 
The idea of the existence of the tree arises when the sentence is 
split up, that is, apart from any word and that is set aside by 
the negative particle. It is said in answer that the negative 
particle can set aside only an idea conveyed by a word. An 
idea which arises apart from any word is not conveyed by the 
word and such an idea cannot be set aside by the negative 
particle. ] 

243. If it be held that the idea (of the existence 
of the tree) is declared false (by the negative parti¬ 
cle and does not disappear) then the negative parti¬ 
cle performs a new function and how can one under¬ 
stand the non-existence of the tree (from the 
sentence Vrkso nasti)? 

[The new function of the negative particle na is to 
declare a certain idea false and not to set it aside. The non¬ 
existence of the tree cannot be understood if this new func¬ 
tion is accepted. Punyaraja points out that the new function 
attributed to the negative particle would be possible only if 
its function is paryudasa and not prasajyaprati$edha. If vrkso ?iasti 
is paryudasa , the negative particle would be connected with 
Vrk$ah and the sentence would mean : not a tree, but some- 


VAKYAPAD1YAM of bhartrhari 

thing similar. If the sentence is prasajyapratifedha, the nega¬ 
tive particle would be connected with the verb asti and the 
meaning would be : ‘the tree does not exist’. It is clear that 
the function ascribed to the negative particle here is possible 
only if the sentence is paryudasa.~\ 

244. If it be said that the negative particle is 
used without reference to any substratum, then it 
could be used even before. If it be said that what 
is negated is the substratum, then the purpose of 
the mention (of the substratum) would be merely 

[ If the negative particle by itself could denote negation 
as well as what is negated, the separate mention of what is 
negated would be for the sake of restriction or elimination 
and not for its own sake. This is an unsatisfactory way of 
construing words.] 

245. Or (the particles) would only suggest a 
restriction or they would restate (meanings expressed 
by other words). Only one word (in a sentence) 
would possess a meaning and the rest would be 

[If the view that the particles ( nipdta ) are only mani- 
festors (of the meanings of other words) and not expressive of 
their own meaning is adopted, then in the present case, in 
vrkfo na the word itself would denote the object and its nega¬ 
tion and the negative particle would only manifest that nega¬ 
tion. What merely manifests the meaning of another word is 
useless. It would do no more than restrict the object to be 
negated to the tree. To interpret a word as merely restrictive 
is not satisfactory. Nor would it do to say that one of the 
two words denotes both the object and its negation and the. 
other word only restates it (anuvada ), because restatement is 
a kind of repetition. All this is the result of looking upon 

II. 245-247 


the sentence and its meaning as divisible. Therefore, it is 
better to look upon it as indivisible.] 

246. In the sentence udahari etc. one under¬ 
stands a contradictory relation between the words. 
But once the whole sentence is finished, a quite 
different meaning is understood. 

[The sentence : udahari ! bhagini \ yd tvam krasdnadvaham 
vahasi sa tvain pracinam kumbham abhidhavantavi adrdksih , referred 
to in the verse, quoted by Punyaraja and the Vrtti is probab¬ 
ly older. Here, once the sentence is finished, a quite diffe¬ 
rent meaning is understood : As the Vrtti puts it— 
vakyasamaptau arthantaropadanam arthantaraparityagaka dr^y ate .] 

247. From sentences the chief meanings of 
which are praise, blame etc. a different meaning 
is understood than the one obtained from the indi¬ 
vidual words. 

[ Sometimes, when the meanings of individual words are 
considered, it consists in condemnation, but the sentence as a 
whole denotes praise. Sometimes, it is vice versa. Punyaraja 
quotes illustrations. In the first one, the verse as a whole is 
meant to praise some king by saying that his glory has whiten¬ 
ed the whole world whereas the parts of the verse enume¬ 
rate four things which continue to remain dark even after the 
king’s glory has spread everywhere. The four things are : 
(1) The spots on the moon, (2) the neck of Siva, (3) Murdri , 
(4) the temples of the elephants of the regions, soiled by the 
flow of their dark liquor. How can one accept the existence 
of these parts of the sentence if they denote the opposite of what 
the whole sentence denotes ? The second verse, as a 
whole, is meant to blame the ocean. But each part of it 
seems to praise it. Another proof that the parts, namely, 
the meanings of the individual words, are unreal.] 

The author now expresses the view of the defender of 
the individual word. 


VAKYAPAD1YAM of bhartrhari 

248. From each word (in a sentence) a mean¬ 
ing not connected (with the meanings of other 
words) is understood which ultimately becomes the 
means of our understanding one single meaning 
from the whole sentence. 

[After understanding the unconnected meanings of the 
individual words, we connect them together and get the 
sentence-meaning which is thus in the nature of connection 
( samsarga ). ] 

This view is now refuted. 

249. The meaning which is first unconnected 
and later joined on to others and thus accumu¬ 
lates becomes something quite different, because it 
is like putting together broken pieces. 

[The translation is according to Punyaraja’s commen- 
tary. His text does not seem to have had na in it. There is 
no na in M. either. It is a pity that the Vrtii on this verse is 
not clear because there are two gaps in the text of it. Though 
T (R) has na in the verse, the commentary Ambakartri explains 
it as though it was not there. ] 

The author now begins the topic of the distinction bet¬ 
ween the primary and secondary meanings of words. 

250. Some thinkers have declared that the 
word which has many meanings is the same word. 
Its power to convey all meanings is differentiated 
according to circumstances. 

[Once the sentence is analysed and we get the individual 
words, consideration of the meaning of the latter begins. 
Twelve views are mentioned :—(1) That the word is the 
same in all its meanings, (2) that it is different with each 
meaning, (3) that the individual word and its meaning are 

II. 250-251 


unreal, (4) that they are real. Each one of these four views 
can be differentiated according to Sabdopacara and the two 
kinds of arthopacara. Sabdopacara means : the application of a 
word to an object primarily, as the application of the word 
go to a cow or figuratively, as the application of the same 
word to a vdhika, because of his resemblance to a cow in dull¬ 
ness. In both these cases, the word go is applied to an object 
which has gotva, really or figuratively. Arthopacara is of two 
kinc j s ;_When the form of the word, whether applied primari¬ 

ly or figuratively, is considered to be its meaning, it is one 
kind. When an outside object is its meaning, whether 
applied primarily or figuratively, it is the second kind. When 
each of the first four views is combined with each of the next 
three views, one gets twelve views in all. But speaking broadly, 
there are two main views, designated as ekafabdadarsamm and 
anekaSabdadarfanam . The present verse begins the consideration 
oftfie former. See, on this topic, my paper on “Bhartrhari on 
the primary and the secondary use of words.” Indian Linguis¬ 
tics, Vol. 29, 1968, pp. 97-112.] 

251 . Therefore, due to purpose or context or 
contact with another word, a word gives up its pow¬ 
er of denoting many meanings at the same time and 
conveys them one by one. 

[ If it is the same word which has many meanings, how is 
it that it does not convey all of them at the same time ? The 
answer is that purpose, context and contact with another 
word determine which meaning is conveyed by the word on 
£ particular occasion. In verses 315-316, othei factors foi 
determining the meaning of a word will be mentioned. Here 
only those three factors are mentioned which help one to 
decide which is the primary meaning and which the secondary 
one. The other factors mentioned later help one to decide 
what is expressed and what is implied. 

The Vrtti says the same thing with examples and adds 
that sometimes words convey more than one meaning at the 
same time: Examples : Sveto dhavati, alambusdnairiydta. See M. 
Bha. I. p. 14, line 14.] 



252. Just as the word ‘cow’ denotes an object 
endowed with dewlap etc. so does it denote a vahika 

[All this is being said according to ekasabdadarsana, the 
view that when a word denotes different things, it is still the 
same word. The word cow denotes a particular class of 
animals through its power called abhidhd. When it is applied 
to a vahika who has not the physical characteristics of a cow 
it denotes that object through the same power and not some 
other. ] 

253. The primariness or secondariness of a word 
having many attributes, that is, the power to denote 
many things, depends upon frequency of usage. 

[Therefore the word gauh primarily denotes the cow 
and secondarily an object like vahika because of more frequent 
usage in the former case. It is the same word which denotes 
both. So this is ekasabdadarsana. 

Commenting on 252 and 253 together, the Vrtti says : 
the same word gauh sometimes denotes the universal as in 
gaur anubandhyah. Sometimes, it denotes the individual or 
substance in which the universal inheres, as in the sentence 
gaur aniyatam or gaur duhyatam. Sometimes it denotes a parti¬ 
cular individual as when one asks a cowherd seated in the midst 
of a herd of cattle: do you see a cow in this herd of cattle? (aslya- 
tra kahcid gam pasyasi )? Sometimes it is applied to a totally diffe¬ 
rent object like vahika because of the presence in him of qualities 
like suffering everything, and eating a lot. Where it is used 
primarily and where secondarily depends upon whether it 
requires the help of other factors like context for conveying its 
meaning. ] 

Another effect of the ekasabdadarsana is now being 

254. One and the same sacred hymn is accept¬ 
ed, without any confusion, as having many mean¬ 
ings and different powers according as it is consider- 



ed from the point of view of the atman , or of the 
gods, or of the ritual. 

[A sacred hymn remains the same even if it is used for 
different purposes such as meditation, muttered prayer and 
sacrifice. On each occasion, it would have a different mean¬ 
ing but the mantra is looked upon as the same. ] 

So far, sabdopacara according to ekaSabdadarSana has 
been shown. In Sabdopacara, the word now denotes one 
meaning and now another. That affects the eternality of the 
relation between the two. So he now considers the question 
from the point of view of arthopacara. 

255. Some hold that due to special circumstan¬ 
ces (in this case the dullness and slowness of the 
vdhlka) the quality of being a covv is traced in a 
vdhlka. There has been a change in the external 
object but the word expresses its own meaning. 

[In this context, two views have been put forward : 
ekasabdadarsana and anekaSabdadarSana. The first view means 
that when a word has many meanings, it is the same word. 
The second view means that it is not but that it becomes as 
many words are there are meanings. In the former view, 
there can be Sabdopacara and arthopacara. When the word 
h is applied to a cow primarily and to a vdhlka secondarily, 
it is the same word. This is Sabdopacara. Within ekaSabdadarSana, 
there can be arthopacara also and that has two aspects : the 
meaning ofa word can be the form itself or an outside object. 
Whether the word gaah is applied to a cow or to a vahika, its 
own form is understood first in any case. Whether applied 
primarily or secondarily, gotva is understood in both applica¬ 
tions. Thus, the form and the meaning are the same in both 
cases. These are the two aspects of arthopacara. As, in Sabdopa- 
cara the word is applied now to one thing and now to an¬ 
other, the eternality of the relation between the world and 
.and its meaning is affected. 


VAKYAPAD1YAM of bhartrhari 

The Vrtti points out that a word is really never used in a 
meaning other than its own. When the word gauh is applied to 
a vahika , it is because some qualities associated with a cow are 
attributed to a vahika : eke$am dcarydndm mukhydt svavifayad 
anyatra §abdasya vrttir nasti. ] 

256. The same form of a word is superimposed 
on all its meanings. It is the object that changes. 
The word is unchangeably fixed to its form. 

[The point mentioned here is that a word never deserts 
its meaning consisting of its external form, namely, a particu¬ 
lar pattern of phonemes. 

The Vrtti attributes to some Acaryas the view that the 
form of a word, specifically its own and internal to it, invari¬ 
ably associated with it and normally never shared with other 
words, is its first meaning. It is superimposed on the object 
which it denotes. The two are identified : Anye tvacarya man - 
yante svarupe Sabdo nityam vartate. Sa eva tasyantarahgo savyabhi - 
cdri $ abdant arena sadhar atio rthah . Tatra cdnupade§ ap r a tip at till 
sarvefdm. rupam tu sabdanam arthe$vadhyaropyate. ] 

The anekasabdadarsaria is now explained. 

257. Those who follow the path of difference 
maintain that though the primary and the secondary 
words are different from each other, they appear to 
be the same because of resemblance in their form. 

[Here Punyaraja points out that those who follow the 
path of difference declare that a word used in a secondary 
meaning is different from the same word used in its primary 
meaning. This is the result of accepting identity as the rela¬ 
tion between a word and its meaning. Once a word has been 
identified with its primary meaning, it cannot again be identifi¬ 
ed with another meaning. One wants a totally different 
word for it. 

A similar question is discussed in the M. Bha on va. 4-10 
and 11-12 in the second ahnika. The question discussed is 


:r~ II. 257-259 

whether a phoneme occurring in different contexts is the same. 
Here the question relates to words having different meanings, j 

258. It is thus that a Vedic hymn for kindling 
the lire becomes different by mere repetition. In 
the same wav, a Vedic mantra that is the object of 
usage or guess also becomes different. 

[ It is prescribed in the Vedas that a certain number of 
mantras should be recited for kindling the sacrificial fire. The 
number is sometimes eleven, sometimes thirteen and some¬ 
times seventeen. Repetition of a mantra is also taught in this 
connection. It follows that a mantra repeated becomes a 
different mantra. It is by repetition that the required number 
is obtained. In the same way, a word used in a secondary 
meaning should be considered to be different from the same 
one used in its primary meaning. 

What is interesting to note here is that the question of 
sameness or difference is thought of in connection with phone¬ 
mes, words and whole mantras. The same phoneme comes in 
different contexts and becomes the object of grammatical 
operations; the same word denotes different meanings in diffe¬ 
rent contexts; the same mantra is repeated in the same context 
and is counted as a different mantra. All this has been said to 
strengthen aiiekaJabdadartana. ] 

2 59. They (the hymns thus repeated) are also 
Veda. For, it is only some that have, been handed 
down. Or rather what is actually handed down is 
meaningless whereas the rest is really subsidiary (to 
the ritual). 

[The mantras which are actually handed down in the 
Vedas are only illustrative. They stand for others also which 
have not been handed down but which are also nonetheless 
Veda. As long as the sequence of the words is the same as 
that of the mantra actually handed down, it is also Veda. Not 
everything is actually handed down. 



Another way of looking at it is that what is actually 
handed down has no meaning other than its own form. So it 
is not mantra. Only that which has a meaning other than its 
form and brings to the mind what is connected with the actual 
ritual is mantra. It is only then that it is subsidiary (Sesa) to 
the ritual.] 

260 . Some maintain that the outward form it¬ 
self is the purpose of the transmission of the Vedas, 
so that everything possessing the form is different 
while those that are used in the ritual are different 
on account of their connection with what is actually 
handed down. 

[If the outward form is the meaning of the Vedas , the 
mantras are different from one another because their outward 
form is different. Not only the mantras which have been 
handed down but the others also have an outward form and 
as that is always different, the mantras are also different. As 
far as the outward form is concerned, a word conveys it 
through Sakti and not through lak$ana. 

261 . The savitrl mantra is different according as 
it is used at the sacrament of initiation or at a Vedic 
ceremony or at low recitation. But it is taken to be 
the same. 

[The savitri mantra used on different occasions is a diffe¬ 
rent mantra , but due to similarity in the sequence of the 
phonemes, it is looked upon as the same. Similarly, according 
to the anekasabdadartana , the word gauh is different according as 
it denotes the cow or the vahika but identity is superimposed 
and so, in regard to the cow, it is said to be expressive ( vacaka ) 
on the basis of prasiddhi and in regard to the vahika , it is said 
to be lak$aka on the basis of aprasiddhi. If there is no super¬ 
imposition, it is a different word and so each is expressive in 
regard to its meaning. Neither is laksaka. To look upon the 
two different words gauh as the same is sabdopacara. 



A mantra, used on different occasions, is a different 
mantra, says the Vrtti. Used in the upanayana ceremony, the 
savitri is called samskarasavitri. When one who has already had 
his upanayana uses it at some other ceremony, it is a different 
savitri altogether. When recited a definite number of times 
as part of an expiation ceremony, it is again different. In 
other words, a mantra differs according to application. It is 
looked upon as the same for practical purposes only. As the 
Vrtti says : vyavaharikam ekatvani]. 

262 . It is on account of the particular form of 
words that they denote their object and so also a 
sentence denotes its meaning by virtue of its own 
form and not of anything else. 

[The Vrtti seems to say here that to look upon the form 
only as the basis for the application of a word to an object is 
cvyutpattipaksa. In the vyutpattipakfa, on the other hand, there 
is some circumstance ( nimitta ) besides the form which is the 
basis of the application : uyutpatti pak;e tu nimittany eva prayo- 

jakani . 

263 . According to those who hold that one and 
the same word has many meanings, the distinction 
between primary and secondary is based on estab¬ 
lished usage or lack of it. 

[Really speaking, the question of primary and secondary 
in regard to meaning cannot arise if the word becomes diffe¬ 
rent with every meaning. As there are as many words as there 
are meanings, each word has one meaning only and that is its 
primary meaning. It has no secondary meaning at all. But 
the question does arise in the ekaSabdadarsana and then long 
usage or the lack of it is the basis for the distinction between 
primary and secondary. 

What the Vrtti wants to say here is not clear as the text 
is doubtful here and there. The point raised is : When a 
word has many meanings, what is the criterion for regarding 
■one as the primary one and the others as secondary ? The verse 



says that the criterion is siddhi or asiddlii. The Vrtti concludes 
thus :— tathaparyayenayasmim vakyc prasiddharthah sabdas tatra 
mukhyah. Anyatra tu gaunah. ] 

264. Others declare that word to be secondary 
which denotes a meaning with the help of the context 
or the presence of another word. 

[ The help of the context or of the presence of another 
word makes the meaning understood from a word secondary. 
This test is said to be common to both ekaSabdadmiana and 
anekasabdadarsana. ] 

The author now wants to express the same idea in the 
words of the author of the Sahgraha. 

265. A word which, when uttered by itself, de¬ 
notes its well-known meaning is the primary one, be¬ 
ing dependent on its own form only. 

[Just before the Verse, the Vrtti says Sangrahakarah 
pathati and then follow 265, 266 and 267 on which there is a 
common Vrtti. Does it mean that all these verses are quota¬ 
tions from the Sangraha ? Punyaraja seems to regard 265 
only as a quotation from the Sangraha. 

266. A word which finds its application through 
the use of another word and by means of a special 
effort as it were is declared by others to be second¬ 
ary, conveying as it does, the secondary meaning. 

[By ’'special effort’, the context is meant here, according 
to Punyaraja.] 

267. In a case where a word takes its chief mea¬ 
ning as the basis for application (to another object) 
the chief meaning is the cause and the secondary one 
the effect. 

II. 267-268 117: 

[That meaning is the primary one to which the word 
can be applied without any impediment or hitch. That mean¬ 
ing is secondary to which the word is applied after some 
incompatibility is felt. When the word go is applied to a 
vahikci , there is first incompatibility but it becomes alright 
when gotva is attributed to the vahika. This is arthopacara. 
This takes place when the word go is supposed to be the same 
in both cases. Though the same, it is looked upon as different 
on the basis of Sciktibhedci and so it is alright. 

The Vrtti here concludes by saying that when a word 
conveys a meaning on the basis of its form only, without 
depending on the context or some other word, it is said to be 
primary in regard to that meaning. If it does depend upon 
these other factors, itis said to be secondary.] 

The author now points out the difficulty in trying to 
take purpose or context as the guide for determining what is 
primary and what is secondary. 

268. The words pur a and drat ; are used in mutu- 
ally conflicting meanings and it is by context that 
we determine what the meaning is on any particular 

[The word purd denotes both the future and the past. 
Similarly drat means both near and far. Only the context 
can tell what the actual meaning is. If the meaning deter¬ 
mined by the context is the secondary one, the meaning of 
these words would become secondary but that would be 
wrong. Both the meanings of these two words are primary. 
Therefore, context is not the correct basis for determining 
which meaning is the primary one and which the secondary 

The text of the Vrtti is doubtful in places but it says 
the same thing and gives other examples besides purd and aiat. 
The point which it makes is that in the case of such words, 
the question of primary and secondary does not arise at all. 
Nor does it arise in the case of pairs like vdyuli vdyuh , aivah 
aSvah , teua tena in which one is a noun and the other is a 
verb, though the two have the same form. Both the meanings 



are primary here also as in the case of pur a and drat :— 
tatha vayur vaytih, atvoivah, tena tena iti ndmdkhyaidndm tulyaru- 
patve visaydntarasyasamsparfdn na gaunamukhyavyavaharo’sti. ] 

How can the question of primary and secondary arise if 
the words and word-meanings are unreal and only the sen¬ 
tence and the sentence-meaning are real ? This question is 
never raised in regard to the sentence-meaning. 

269. Once the analysis of the meanings of the 
individual words from that of the sentence is done, 
the meaning of one word so obtained can be connect¬ 
ed with another, also so obtained. 

[When the sentence and its meaning are considered 
indivisible, there are no individual words and their meanings 
and so the question of primary and secondary cannot arise. 
And yet for practical purposes the sentence is split up into 
words by a process of abstraction ( apoddhara ). When this is 
done, the distinction of primary and secondary is based on 
long usage or lack of it. ] 

But then a word is sometimes used without being linked 
to another word. That means that the individual word is real. 

270. When a word is sometimes used by itself 
with the verb asti (understood), it is really a sentence. 
That is why it is not connected with any other word. 

[What is meant here is that the verb ‘tobe’ is understood 
and added on mentally to a single word when it is used. Thus, 
it is really a sentence and so no other word is connected with 

271. When the word ‘cow’ or ‘horse’ is uttered 
as an answer to the question ko’ yam? (what is this?) 
there is some action like seeing hidden in the question 

[The question K’yam ? = What is this ? really means : 
What is this that is seen ? =Ko\ym drsyate ? in which the action 

II. 271 273 


of seeing is understood. Or it may be some other suitable 
action. In the same way, the answer ‘cow’ reallymeans: ‘it is 
the cow which is seen’ or some other suitable action is implied 
in it. An action is implied both in the question and in the 

[The Vrtti also points out that a single word is never 
used. Whether it is a question or an answer, it always 
implies an action :—. . Mi kriydvisesam abhyantarikrtya praina- 
pralivacane bhavatah. If the question implies an action though 
it is not openly mentioned, it is natural that the answer also 
should imply it :— tatra ca prainakala evantaritayam kriyayam 
prativacane’pi tadvifayah caritakriyah Sabda upadiyate. ]. 

Some have taken the presence of the original properties 
in a greater or less degree as the basis of the distinction bet¬ 
ween primary and secondary. 

272 . Nor can abundance or deficiency in the at¬ 
tributes be the basis of the distinction because some 
consider even abundance to be deficiency due to 
long usage. 

[Dullness is associated with the cow. When the word 
cow is applied to the animal, it is a case of primary application 
because, in the cow, dullness is found in a greater degree. 
When the same word is applied to a vahika, it is a case of 
secondary application because dullness is found there in a 
lesser degree. But this criterion is unreliable as abundance 
and deficiency are relative terms. The vahika may be really 
duller but long usage considers that dullness exists in him in a 
lesser degree. 

The Vrtti discusses, besides the above example, simho 
manavakah also :— sampurnaka saktyadayah simhe, nyund mdna- 
vake. gavi ca sampurnajadyadayo nyilnastu vahike. 

The author now speaks about similarity as the basis of 
the distinction between the primary and the secondary. 

273 . When a word expressive of the universal is 
applied to something in which the universal is absent 


VAKYAPADlYAM of bhartrhari 

but a similar attribute is present, that word is said 
to be used in a secondary sense. 

[According to thisview, the word cow denotes the uni¬ 
versal ‘cowness ’ (gotva). It exists only in the cow. But when 
the word is applied to a vdliika, it is not because there is gotva 
in him, but because he resembles the cow in being dull. 
Resemblance in some attribute is, therefore, the cause & of the 
secondary application of the word. 

As the Vrtti puts it :—taira rudhasambandhasya prasiddliasa- 
hacaryasya dharmdnlarasya darsandd atajjdtlyefu jatitabdal, prayu- gauna ityucyate. ] 

The author now speaks about viparydsa, misapprehension 
on the basis of the distinction between primary and secondary. 

274. When an object appears like another as 
though by misapprehension, then words like cow ex¬ 
pressing the latter are said to be in a secondary sense. 

[One can apply a word expressive of one thing to an¬ 
other, either through misapprehension or deliberate superim¬ 
position. To call mother-of-pearl ( sukti) silver after misappre¬ 
hension in the dark is an example of the first case. To apply 
the word cow to a vahika on seeing some resemblance between 
the two and not. because of misapprehension is an example of 
the second case. Here there is no wrong identification be¬ 
cause one clearly sees the difference. That is why the word 
ivais used after viparydsa. ] 

The author now speaks about shape and power as the 
basis of the distinction between primary and secondary. 

275. Just as a,v plough or a sword or a pestle, be¬ 
ing endowed with a particular shape and power, is 
understood as the fixed accessory ( sadhana ) of a 
particular action, 

276. (Just as )these things have no power connect¬ 
ing them with other actions but are invariably under- 


II. 276-278 

stood as meant for the performance of particular 
actions because of their shape, 

277. In the same way, a word endowed with a 
form and power from the very beginning has its 
purpose fixed. Through some power, it can also be 
applied to other things. 

[ These instruments and tools perform their particular 
actions on account of the particular shape with which they 
are endowed. One cannot cut grass with a plough nor fight 
a battle with it. In the same way, the power of a word to 
denote its meaning is fixed from eternity. This is the view 
not only of the grammarians but also of the Mimamsakas. See 
Jaimini, Mi. Su. I. 1.5. Also Vak. III. Sambandha. 29.] 

278. The meaning which is understood to be 
the purpose (that is, the main one) of a word as 
soon as it is heard is considered to be the mainmean- 
j n cr whereas the one to which it is applied by a 
special effort is the secondary one. 

[ If a word like go has many powers and it can be appli¬ 
ed to a vdhika also, how is one to decide which is the primary 
meaning and which the secondary one ? The answer is : that 
which is understood directly from the word without the help 
of the context, by merely hearing the word is the main one. 

That which is understoodwith the help of the context and 
otherwords in the sentence is the secondary meaning. 

Commenting on 275-278 together, the Vrtti points out 
that certain tools and instruments, by virtue of their shape 
and power, are associated in the world with particular 
actions. Not that they cannot be used for doing other 
actions, but normally, they are meant for particular actions. 
They are not associated with nor called by the other actions : 
loke te tu karmantaresupadiycimanah sadhayanto’pi tamartham 
tadangatvena vypadesam na labhante. Similarly, words have 
normally the power to convey particular meanings, but they 
are sometimes used to convey other meanings secondarily but 


VAKYAPAD1YAM of bhartrhari 

they do not become known in the world as being expressive of 
these meanings : pratyayanaSaktimairam tupadayanyatra prayujya- 
mdndyah pratydyayanto'pi tanarthanstatsambandhitvena loke 
vyavasthayd prasiddhim na labhante.'] 

279. When the words go, Yusmad and mahat have 
the suffix cvi at the end and are used in a meaning 
other than their own, what was not so becomes so 
secondarily and sometimes also primarily. 

[In the expressions : agaur gauh sampadyate, go'bliavat, 
atvam tvam sampadyate, tvad bhavati, the quality of being a cow 
and of being ‘thou’ is superimposed on what is not a cow and 
what is not thou. What is not a cow does not really become a cow. 
It is onlya superimposition. In amahdn mahan bluitah, mahadbhu- 
tas candramah, what was not big does actually become big. This 
case is therefore different from the two previous cases. The 
word mahat is used in its primary meaning in mahadbhutah, but 
in go’bhavat, go is used in a secondary sense, because it is 
applied to something which does not really become a cow. 
Being used in a secondary sense, the o in go is not considered 
to be pragrhya. For the same reason, there is no second person 
suffix in tvadbhavati. There is no real yusmad here. It only 
exists secondarily. In the words of the Vrtti—agaur gaur 
abhavat, go’bhavat iti pragrhyasarhjna. na bhavati. Atvam tvam 
sampadyate, tvad bhavati madhyamo na bhavati. ] 

But, says the objector, this is alright in such expressions 
as go'bhavat and tvadbhavati. But in mahadbhutas candramah, the 
use cannot be called a secondary one as the moon does really 
become big {mahat) on full-moon day. That being so, the 
long a, taught in P. 6.3.46, at the end of mahat should come 
in. To show that even in mahadbhutah candramah, the use is a 
secondary one, the author says— 

280. The original state is changed into m agnitude 
or whiteness. Considered as something different, 
it (the original state) is the cause of secondariness. 

II. 280-282 


[ In mahadbhutas candramah = ‘the moon has become big’, 
and Suktibhavati patafi = ‘the cloth has become white’ also, 
there is secondary usage. The moon remains the same, but 
it is thought of as having two different states. The former 
state is considered as the original one and the later state as 
the modified or secondary one. It is a question of the speakers’ 
intention. When the former state is thought of as the original 
'ne ' the later state appears as the superimposed one and then 
’ j arv usage results. On account of this secondary usage, 
the final of mahat does not become a by P. 6.3.46. The pre¬ 
sence of secondary usage here also is stated by the Vrtti as 

f UoWS __ purvottarayor avastliayor aSritayoh savyaparatvat purva- 
sya avast hay a vivakfayam satyam vikaraiabdasya tadupagrahi gauna- 
tvam vijnayate. ] 

281 The proper nouns Agni, Soma, etc., which 
are united with their names on the basis of their ex¬ 
ternal forms are really used in a secondaiy meaning 
in phrases like Agnisomau manavakau ) because these 
words (as applied to manavakau) are not well-known. 

[Words like agni, and soma are names of deities which 
flr _ their primary meaning. When they are applied to human 
, • s on the basis of some resemblance such as brightness, it 

be T S case of secondary usage based on superimposition. When 
thev are given as names to new-born individuals, they are 
ajn used secondarily, but this secondary usage is not based 
on resemblance but on the mere fact that these words are not 
known as the names of new-born infants. In other words, 
abrasiddhi is the basis. That is why there is no sattva in Agni¬ 
somau according to P. 8.3.82.] 

It is now stated that when there is superimposition, the 

usage is, of course, secondary. 

282. When Agnidatta is called Agni, the word 
makes its primary meaning (of god Agni) subordi¬ 
nate and, therefore, the usage becomes secondary, 
as it also stands for datta. 


VAKYAPAD1YAM of bhartrhari 

[ When Agnidatta is referred to by the word Agni, the 
primary meaning of the latter word, namely, the god Agni, is 
made subordinate and it now denotes a person called Agni¬ 
datta by superimposition ( adhyaropa ) . This is also secondary 
usage. ] 

It is now explained how the augment sut in HariS- 
candra is justified even when the word is only the name of a 
student ( manavaka ). 

283. In the derivation of words which is based 
on different circumstances, the presence or absence 
of sut in such words as Hariscandra are regulated. 

[There is the augment sut in the word Hariscandra and 
it is correct when it is the name of a r$i (P. 6.1.153.) When 
we make a compound word in the sense of HariScandroyasya, 
the resulting form would be Haricandra , that is, there will be 
no sut. The circumstance being different, the form is also 

The Vrtti points out that sometimes the form with sut 

can be correct when it is the name of the Rsi :_ tatha sati 

Rser anyatrapi sasutkasya sadhutvam vijnayate . ] 

284. A word which attains its correct form as 
the name of a jRsi and is then applied to denote 
something else does not lose its inner correctness of 
form when applied to a different object. 

[A word gets its correct form before it enters into a 
sentence and ibkeeps that form even after entering into the 
sentence in a secondary sense. As the Vrtti puts it . . .tadd 
svavifaye siddhah bahirangdrthapradurbhdve na nivartate .] 

Some deny the distinction between primary and second¬ 
ary meanings. They are answered as follows— 

285. Whenever a meaning, however contradic¬ 
tory it may be, is understood from a word,then accor¬ 
ding to that understanding, that is its primary 

II. 285-287 


[ Whatever meaning figures in the mind when a word or 
sentence is heard is its meaning, however unusual or contra¬ 
dictory it may be. This is based on the Adimmnsaka principle, 
expressed in the following Mi. Su : —Sabdavattupalabhyate 

t add game hi ddyatc tasya jiianamyatha? nye$am . (Mi. Su IV 1.6.15) 
The Vrtti gives the example of the mirage. The mirage 
looks like water’from a distance. So it is mistaken for water 
and the word water applied to it. This application should be 
looked upon as a primary one because it is based upon the 
well-established meaning for the > application of the word 
water : —jalaiiirbhasdydm lii mrgatrpiikdydm buddharutpannayam 
mukhya eva jalaSabdah prayoktavya iti. Tulyam hi pravrttinimittam 
sarvatra sabdasya prayojanam bhavitum arhali. ] 

It is now shown that even if the meaning is what 
figures in the mind, there can be a distinction between what 

is primary and what is secondary. To show this, the author 
says something about what is real and what is not. 

286. Even though the determination of an object 
depends upon our understanding of it, still all un¬ 
derstanding of it does not go in vain (or, all under¬ 
standing of it is not as in the case of the well-estab¬ 
lished object, if the correct text is prasiddlia iva, in¬ 
stead of asiddhaiva.) 

[ The Vrtti seems to point out that even though the 
determination of an object depends upon our understanding 
of it still our subsequent experience in regard to such ob¬ 
jects’ brings out a distinction between the primary and the 
secondary :— ‘pratyayadhine ’ pyarthasyavadharane kvacit tadvisaya- 
nam pratyayandm avyabhiedrena yd pravrttir loke ( sa ?) gauna-mukhya- 
bhavam vyavasthapayati .] 

287. The perception of (real) water and of such 
things as a mirage is the same. In spite of the simi¬ 
larity of perception, mirage is not water. 

[ The point here is that mere cognition is not enough 



for determining the existence of an object. We take a mirage 
to be water, but there is no water there. 

The Vrtti points out that our perception of real water 
and our perception of mirage as water resemble each other. 
But the difference becomes clear when we go to the place and 
try to touch the water in the mirage, drink it and have a bath 
in it. So the use of the word ‘water’ to the water which we 
see in a mirage is only secondary usage. As the Vrtti puts it— 
tathapi spars ana-snana-panadinam abhavdt taddesapraptau 

cadarsanan nedam salilam iti . . . tasu mrgatr$nikdsu nasti rnukhya- 
sya sabdasya pravrttih.] 

288. Even though there is similarity in the per¬ 
ception of a serpent and that of a rope, still their 
difference is obvious from their well known special 
activities (that of biting in the case of a snake and 
that of binding in the case of a rope. 

[We can conclude that the objection is non-existent if 
we do not see its special activity, even though it may figure 
in the cognition. The snake may figure in our cognition of 
the rope, but as the rope does not bite, we can conclude that 
it is not a snake. Further, that our perception ofit was wrong. ] 

289. Whatever difference is perceived as a result 
of any circumstance causing a contrary perception, 
that they call a false perception. 

[A defect in the senses or in the object itself may cause 
a false perception. What is seen because of such a defect is 
said to be false. 

The Vrtti mentions the following as possible causes of 
wrong perception :— Santamasa timiropaghata , madya , visapdna , 
anyadesavasthana. When, due to these causes, the wrong per¬ 
ception takes place, it can be corrected by comparison with 
perception under normal conditions and by verification 
through touch etc.— tattu spar sanadibhir yathabhutamavadhdry ate. 
timiropaghatad darsanabhede dvitiyasya vastunah sparSabhavada- 
paricihinnacaksu$o yd pratipattih saiva nydyyetyavasiyate. ] 

II. 289-293 


More examples follow to show that two cognitions may 
be similar if their objects are similar and refer to their special 
property but that they are really different. 

290. Even though one sees ups and downs in a 
picture, similar to those of mountains (nimnonmtGm 
citre) still it does not cause obstruction etc. (as a real 

[ R and RP have nimnonnate citre but nimnonnatam is a 
better reading. The Vrtti and the Karikd text of M. have it.] 

291. Just as it is possible to have continuous 
contact of the hand with the wheel, that is not possi¬ 
ble in the case of the torch-wheel which can be touch¬ 
ed only with interruption. 

2 92. While it is possible to have contact and 
protection in the case of cities by means of forts, 
walls and turrets, it is not possible to have these in 
the case of imaginary cities. 

2 93. As much work as it is possible to get from 
real animals cannot be got from earthenware imag¬ 
es of them. That is why the affix Kan has been 
taught by Panini after (words expressive of imi¬ 
tations ). 

[ It has been shown so far that the distinction between 
primary and secondary on which some grammatical opera¬ 
tions are based depends upon the reality or unreality of the 
objects denoted by the words in question. One can tell 
whether the objects are real or not from the nature of their 
cognition which, though similar, may be different because of 
difference in their nature. Now the author speaks about 
difference between primary and secondary words, the basis 
for the use of thesuffix Kan. P. 5.3.96. teaches the suffix Kan 
after words expressive of images of objects. One thus gets 

128 VAKYAPADlYAM of bhartrhari 

forms like asvakt z, ns It ole a , gardabhaka, As the objecls'are not 
real animals but ouly their images, the suffix Kan can be add¬ 
ed. Mere resemblance is not enough. It must be an arti¬ 
ficial image. There is, of course, secondary usage here also. 
As the Vrtti puts it Tasmat. kasyacid eva sadriasya dharmasya 
bhdvadupamanopameyablmvasambandhe sati ‘ivepratikrtau’ (P 5.3. 
96). ityetatprakaranavihitanam pratyaydndmutpattau nimittam 
labhanle. ] 

294. Mountains and other well known objects 
cover a wide area but their reflection covers only a 
small area. 

[The use of the word Parvata in the sense of the reflection 
of a mountain is secondary usage and so it takes the suffixfan 
and we get the form parvataka. To call an earthenware image 
of a horse asvaka is secondary usage based on similarity of 
shape. To call a reflection of a mountain parvataka is also 
secondary usage based on resemblance but resemblance due to 
the relation of original and reflection. ]. 

295. While real poison and other such things 
are the cause of death, the same things seen in a 
dream are not capable of causing the same. 

[The implication is that the suffix Kan can be added to 
words expressive of things seen in a dream, because they do 
not have the effect which the same things have in real life. 

The Vrtti says :—Vi ? abhojanddayo hiyathd prasiddhd marana- 
ksutp ip as dp ratighatahetutvena vyavasthitd loke na tathd svapnonmSda- 
murcchadisu. tasmanna te mukhyavisayatvam labhante']. 

296. Things that seem to be otherwise as a 
result of changes relating to time, place or the senses 
are, however, understood rightly by following world¬ 
ly tradition. 

[ Whether a thing is distant or near is due te difference in 
place. The rays of the sun shining in summer and thus causin° r 




sing the illusion of water in a desert is due to time or reason. 
One who suffers from an eye-defect sees two moons instead of 
one. This is an instance of an object looking different due to a 
defect in the senses. 

The Vrtti gives other examples which unfortunately are 
not clear due to gaps in the text. ] 

297 . The world does not regulate its verbal 
usage on the basis of knowledge arising from defec¬ 
tive senses or which is of a supernatural character. 
Words are based on worldly tradition. 

[ Knowledge due to some defect in the senses is the 
opposite of truth and is illustrated by mirages and the seeing 
of two moons. Supernatural knowledge is that of Rfis. 
Neither of them is fit for worldly transactions. 

Though the Vrtti has gaps here also, the following rele¬ 
vant sentences may be noted—. . .yoginam sarvajnandm ca 
jnanam SabdavyavahdrefU tair apt ndnugamyate. Prakrtalokadrfti- 
nibandhanatvacca Sabdarthasambandhasya sarvenarthabhidhdne yat- 
nam kurvata lokahprathito nugantavyah.] 

298. Just as a lamp reveals, in an object like 
a jar, through association (or proximity) other 
things than that for the illumination of which it was 


299 . In the same way, a word conveys, from 
among the things which are connected together, 
those that are different from the one to convey 
which it was used. 

[When a lamp is used to illuminate an object, it illuminates, 
not only that object but whatever else is invariably associated 
with it. If a lamp is lighted to see ajar, one sees not only the 
jar but also the properties which are inherent in the jar. This is 
inevitable. Thejar itselfis the main thing, its properties are the 
invariably associated things. In the same way, when a word 
is used to convey a particular property from among the many 



which are connected with one another, it conveys not only 
that property but also others which are always associated with 
it. Punyaraja takes the word bhava as an example. On the basis 
of its derivation, it can convey several connected notions such 
as the fact of being something finished ( siddhatva) , the mas¬ 
culine gender, singular number and so on. If it is used for 
conveying the idea of being something finished, it will not stop 
there. It will also convey the masculine gender, singular 
number etc. 

The illustration of the lamp and of the -word bhava go back to 
the Vrtti. The former, of course, is mentioned in the verse298 it¬ 
self. The Vrtti contains several references to the Mahabhasya on the 
sutra: Bhave (P. 3.3.18.) 

The lamp and the word bhava were jfidpaka examples. The 
author now takes a karaka example. 

300. Though the churning of the ignition 
sticks ( arani ) is done for producing fire, it also pro¬ 
duces the unintended smoke in the same process. 

301. In the same way, a word also, when a 
particular meaning is meant to be conveyed, denotes 
by association, an unintended meaning also. 

[The Vrtti points out that neither the unintended smoke 
which is produced nor the unintended meaning which is con¬ 
veyed plays any part in the action :■ —aprayojakatvdttu pratito’- 
pyarthah kriyasadhanabhdvena na parigrhyate. agniprayuktena hi 
nirmanthanenaikasadhano pi dhumasannipatah pdkddi$u ca drs ta~ 
phale$varthe?u agnivannopadiyate. 

That a lamp should illuminate not only the intended ob¬ 
ject but also what is next to it is natural. A word, on the other 
hand, is used to convey a particular object. Why should it 
■convey something more than that?] 

This objection is now answered. 

302. Just as one cannot abandon (while taking 
a thing) something which is very closely connected 
with it, in the same way a word cannot but denote 

II. 303-305 


•what is intimately connected with its primary 

[ Punyaraja, following the Vrtti explains that this is like 
one’s not being able to avoid bones when one buys fish or meat. 
Similarly, a word not only conveys its primary meaning but 
also its gender and number— tathaikavacanadir api sabdo vinapi 
tena pratipadikarthasydpratyayakatvadairitasairisargah tyaktum na 

303. Even when the unintended meanings are 
present and they are conveyed, it is still the 
meaning which led to the use of the word that is the 
accepted one, even though there is no difference in 
the form of the word (with which the others are also 


[The word which conveys the main meaning, that is, the one 
which led to its use is the very word without any difference in 
form which conveys the other meanings also. 

The Vrtti points out that even though it is the same word 
without any change in its form that conveys all the meanings, 
it is always possible to distinguish between the one which leads 
to the use of the word on a particular occasion and the 

others’_-. .tulyaSrutir dpi sabdasteuaiva rupena piavartamdiia iha 

viiaye pr ay oj akenar thcndr thavdn ihaprayojakeneti samarthyad avadha- 
rayiturn §akyate. ] 

304. Sometimes the meanings are not meant to 
be primary or secondary, sometimes even proximity 
is not the cause of a meaning being conveyed. 

305. Sometimes a meaning not belonging to the 
word actually used is conveyed and sometimes it is 
the chief meaning which conveys another. 

[ Even when the meaning which is the cause of the appli¬ 
cation of the word is the expressed meaning and not what 
is necessarily understood, there are four possibilities: 



(1) reversal of what is primary and what is secondary, (2) part 
of a meaning is not meant, (3) the whole of the meaning is not 
meant, (4) the conveying of another meaning without giving 
up the first meaning. These possibilities are now explained one 
by one. ] 

The following is an example of the reversal of what is 
important and what is secondary. 

306. ( In the sutras setting forth the meanings 

of words ending in taddhita suffixes) the verb just says 
something about the meaning of the taddhita suffixes 
(and does not insist on action as it usually does) 
One sees a reversal of the relation of primary and 
secondary meanings. 

[The following is an example. P. 4.4.2. teaches the 
meaning in which certain suffixes are added to stems. As the 
verb is the most important word in the sentence and as a verb 
like divyati primarily expresses action, the word ending in the 
suffix concerned should primarily express action. But, in fact 
the usual relation of primary and secondary between action 
and accessory is not meant here. In divyati, the action is 
primary but in akfika the word formed according to that sidra 
P. 4.4.2., (akfair divyati =dkfikah) , the agent is primary and not 
action. In the suttra Bhdve (P. 3.3.18) the masculine gender 
and the singular number are not significant. 

While Punyaraja explains this verse on the basis of P 
4.4.2, the Vrtti does it on the basis of P.4.2.59 and P 5 2 84 
where also the meaning of the suffix is indicated by means of a 
verb. The word formed by the suffix concerned, however, expres¬ 
ses the agent primarily and action secondarily : sarvadhdtnpa- 
sarjanibhutakriyaparicchinnam dkhydtdpratyavamrs taruparriprddhdnyena 
sadhanam abhidhiyate . ] 

Now follows an illustration of a part of the meaning 
being not meant:— 

307. (ab) Proximity is not the cause of the 
gender and number being conveyed in the sutras 
teaching meanings of suffixes. 

II. 307-309 


[ In the siitra Blidve (P. 3.3.18). the masculine gender 
and the singular number are not significant. Words formed 
by the suffixes concerned can be in other genders and numbers 
like paktih, gamanam, pakau, pdkdh, etc. ] 

30 7. (cd) From the word hrasva= short (in P.1.2. 
32.) it is the unmentioned measure which is under¬ 

[ In P. 1.2.32. the word ardhahrasva means half a mdtrd. 

The Vrtti points out that in P. 3.3.18 the word bhava 
cannot convey its meaning without some case-ending or other 
which brings about the correctness ( Sabdasamskara ) of the word 
which is its only purpose and not to convey the masculine 
gender and the singular number as significant items ; bhava 
ityevamadifti svarthasydnyatlia vaktum aSakyatvdc chabdasamskdra- 
nimittalvdcca sannihite pi lihga- samkhye pratyaydnte samuddye vacya- 
tvam na pratipadyete. Similarly in ardhahrasva in P. 1.2.32, the 
word hrasva has no significance. It means one mdtrd and 
nothing more. Ardhahrasva, therefore, means half a matra and 
nothing more:— na hyatra hrasvendsannihitena vd prayojanam 
kiilcit. . . . hrasvasyardhopalak$anatvenopadanam. ] 

308. Half of a short syllable is really what is 
meant even though the short syllable itself may 
not be mentioned because short is meant to include 
everything that has the duration of one matra. 

309. The words ‘long’ and ‘protracted’ may 
qualify ‘half’ (so that it means half of the mdtrd 
belonging to the long or the protracted) or mdtrd 
may quality ‘half’ so that the whole means half a 
mdtrd. Or the word ardhahrasva may refer figuratively 
to the universal of half a mdtrd just like saptaparna. 

[ The word ardhahrasva in P. 1.3.32 may be understood 
in three ways : (1) hrasva means mdtrd, so ardhahrasva means 
half a mdtrd. (2) hrasva stands for all the three ; hrasva, dirgha, 



and pluta and ardha means a part, equal or unequal When 
added to hrasva , it would stand for equal part, because the 
total duration of hrasva is one matra and half a matra would be 
an equal division of one matra . As hrasva means half a matra , 
when added to dirgha or pluta , it would mean an unequal part, 
the other part in dirgha and pluta being 1£ and 2\ mdtras 
respectively, (3) the word ardhahrasva is indivisible and it 
stands for the universal (jati i of half a matra. 

The above interpretations are all found in the Vrtti as 
the following relevant sentence shows —sa hi tadavayavopalak$a - 
natvad dirghacaturbhaga iti vocyate , plutasadbhaga iti vd mdtrarddham 
iti vd. Tat sarvatha tasya parimdndvasthd vyakhydtd bhavati.] 

It is now shown that, sometimes, it is the chief meaning 
that indicates another. 

310. ‘One has to go, look at the sun 5 though 
this sentence says something about the sun, it really 
indicates time, because it says, with a hint at the 
means ( upaya ) that time should be ascertained. 

[The main meaning of the sentence relates to the sun 
but that is only a means to an end which is the ascertainment 
of time. ] 

311. In the sentence vidhyaty adhanusd ‘he 
pierces with something else than a bow 5 the parti¬ 
cular weapon stands for weapon in general or for 
anything that can be the substratum of the power of 
being a weapon. 

[This is an example of the particular standing for the 
general taken from the vyakaranaSastra. It is an example of 
kvacit pradhanam evartho bhavatyanyasya laksanam (v. 305), be¬ 
cause the word dhanuh , without giving up its main meaning 
‘bow’ stands for any weapon. 

The Vrtti is not clear because of gaps in the text but it 
seems to say that in P. 4.4.83, the word adhanusd is a paryudasa 
kind of negation and means ‘something else than a bow’. But 

II. 312-314 


in this verse it is taken as standing for weapon in general. 
The idea is that by negating a particular weapon, weapon in 
general is conveyed. ] 

A similar example from the world is now given. 

312. When a boy is told : ‘save the butter from 
the crows’, he does not refrain from protecting it 
from dogs etc. knowing that the order refers to des¬ 
tructive agents in general. 

The Vrtti says:— Kakebhyah kakdd va sarpih samraksyatcm 
ityukte upaghdtahetusamdnyamdtropalaksanatvdt kakajdter ydvan 
upaghdtahetuh sa bratiyate. By saying that kdka in the 
verse stands for any destructive agent ydvan upaghdtahetuh, the 
Vrtti makes it clear that in worldly verbal usage also, the 
particular stands for the general.] 

313 . ( When an order for feeding somebody is 
given) the washing of the dishes and plates, though 
not actually mentioned, is also understood because 
it is part of the action of feeding. 

[The point made here, is that what is invariably concomi¬ 
tant with the chief meaning is also conveyed even though no 
word expressive of it is used.] 

The factors which enable us to decide which meaning is 
primary, which secondary and which implied are now enume¬ 

314 The meanings of words are determined 
according to the sentence, situation, meaning, pro¬ 
priety, place and time and not according to mere 
external form. 

[The following examples are given by Punyaraja. By sen¬ 
tence is meant the fact of construing together several words 
occurring in the same sentence or taking together several 
sentences occurring in different contexts. For example: in the 



sentence katam karoti b hism am udasam darsaniyam , the connec¬ 
tion of the verb karoti is with the word katam but as there cannot 
be a substance without qualities and as qualities must have a 
substratum in which to inhere, the other words like bhismam 
ending in the second case-ending are taken as qualifying 
katam and so a mat having those qualities is made and not any 
other on the basis of interconnection within the sentence, 
when someone is asked to fetch saindhava in the context ( praka - 
rana) of battle, he would fetch a horse but when the same 
word is uttered in the context of eating, he would fetch salt. 
Thus context helps to determine the meaning of a word. When 
the same word is used in different sentences, its meaning is 
determined by the meaning ( artha ) of the other words in the 
sentence as in the following:— anjalina juhoti, aiijalind suryam 
upatifthate, anjalina purnapatram aharati. In these three sentences 
anjalina means something different due to its association 
with the meanings of the other words. It is propriety ( aucitya) 
that helps to determine, when sira, asi and musala are mention¬ 
ed without reference to any particular action, that sira is for 
ploughing, asi for fighting and musala for unhusking again. 
When one says that he comes from a town to the east of Math¬ 
ura, from the word ‘to the east’, one would understand that he 
comes from Pataliputra. When one only says ‘the door’ in 
winter ‘the listener’ would understand that it has to be closed 
and when the same thing is said in summer, he would under¬ 
stand that it has to be opened. Here it is time or season which 
helps to determine the meaning. 

In the Vrtti, some examples are the same while the others 
are different. But the text being corrupt, the meaning is not 
clear. ] 

The author now gives two verses giving a more complete 
list of the factors which help to determine the meaning of 
words. These may be quotations from some unknown work. 

315. Connection, separation, association, oppo¬ 
sition, meaning, context, indication, the presence 
of another word, 

II. 316 


316. Suitability, propriety, place, time, gender 
and accent etc, these are the causes of our determin¬ 
ing the meaning of a word when there is no definite¬ 
ness in it. 

[The Vrtti and Punyaraja record that it is always samarthya 
= ability or capacity of the word which helps to determine its 
meaning when there is indefiniteness. The so called factors 
enumerated only reveal this capacity : Tatra kecit sdmarthyame - 
vaikam bhedavadharananimittam iti many ante . yasydrthaprakara - 
ndbhydm svdbhaviko blicdah samad bigamy ate so'pi samarthyad eva. 
samar thy am hi bahudhd pravibhajyate. (Vrtti). 

As for the examples of the different facts, the Vrtti and 
Punyaraja give the same ones mostly. It is obvious that the 
latter has taken them from the former. Here and there the 
Vrtti gives examples which are ignored by Punyaraja. It also 
raises some points for discussion which are ignored by Punya¬ 
raja. It is anxious to give an idea of the variety of views 
which existed in its day on some of the points raised. That is 
why the expressions : anye, anye tu , anyetvahuh , kecit, keciddhuh , 
ye$am, te$am , ke$am-cit occur in the Vrtti in v. 315 and 316. 
The examples common to both are as follows— Samsarga = 
connection:— sakiSora dhenur dniyaldm. Here on account of 
the connection between dhcnuh and ki$ora> a mare is understood 
from the former word and not a cow. Viprayoga= separation— 
akiSora dhenur aniyatam = c let the dhenu without its ktiora be 
brought. 5 Here also, by dhenu a mare is understood because 
it is only a mare which can be without kiiora . Sahacarya 
(association) — Rdmalak$maiiau. Because of association with 
Lakmana , Rama here means son of Dasaratha, Virodha 
(opposition) — Ramdrjunau. Here Rama means ParaSurama 
because of his well-known opposition to Arjuna. Artha 
= meaning and prakarana — context have been illustrated 
under the previous verse. Linga — indication— aktdh Sarkardh 
= sugar mixed with grease. From indications found elsewhere 
it is understood that butter is used for mixing. Ramdrjunau 
can be taken as the example of the determination of meaning 
through the presence of another word. Samarthya^ suitability, 
abhirupaya kanyd deya= the girl must be given in marriage to a 


VAKYAPAD1YAM of bhartrhari 

good-looking person. Through suitability, one understands a 
good-looking bridegroom. Auciti = propriety, desa~ place, kala = 
time have already been illustrated under v. 314. It is well 
known that in the Veda, the meaning of a word is often deter¬ 
mined by its accent ( svara ).] 

317. Even, if it is held that a word having many 
meanings is different in each case ( bhedapakse ), the 
external form being similar, confusion arises and 
it is with the help of context etc. that difiniteness 
in meaning is arrived at. 

[Whether one holds the nandtvapakfa or the ekatvapakfa, one 
has to take the help of the context in determining the mean¬ 
ings of words having more than one meaning. If the word is 
different with each meaning, the form being the same, one 
cannot decide which meaning to take and the help of the 
context is necessary. If the view is that it is the same word, 
the help of the context is all the more necessary.] 

Sometimes a word has more than one meaning, because 
from its form only, it can be a verb or a noun. Then also 
context alone can determine the meaning. 

318. When words have the same form as nouns 
and as verbs and are to express a different mean¬ 
ing as each, then the meaning cannot be understood 
from the external form only. 

[The word a$va as a noun means a horse but it can also be 
a verb. From the root tu o tvi gativrddhyoh one would get the 
form a$va in the aorist second person singular ( luhmadhyama - 
puru$aikavacana). In the same way, ajapayah can mean goat’s 
milk or the word can be the causative form of the root ji = 
to conquer. In such cases, only the context can be the 
guide. ] 

319. An enlightened hearer knows that praise 
and blame, meant to promote action and abstention 
from action respectively, are really unreal. 

II. 320-324 


320. The praise of a prescribed action having a 
visible or invisible fruit only serves to prompt the 
agent thereof. 

[These verses are meant to show that the individual word 
and its meaning do not really exist. Only the meaning of the 
sentence is real. The chief meaning of a sentence is action and 
though a part of it may consist of praise of this action, it has 
no real existence. In the same way, the condemnation of an 
action contained in a sentence meant to keep people away 
from it has no real existence, that sentence being meant only 
to keep people away from that action. 

The Vrtti gives the following example of praise— brahmacary- 
enaivehanantam dyur avdpnoti, ghrtena pdfimdnam apahantiti. tatra 
brahmavraiam caritavyam, ghrtam pdntyam ityetdvdn upadesah.] 

32 1 - Tust as a crying child is put off when he is 
threatened that a tiger would eat him, in the same 
way, some bad consequence though unreal, is held 
up (before one who does some prohibited act). 

322 . Therefore, a learned man does not undei- 
take an action prohibited by the sastra after having 
arranged for averting its bad consequence. 

32 3. One should not violate the injunction that 
‘one should not touch it (the sacrificial cake) with 
the teeth while eating it, for, if one does so, serpents 
would destroy him’ after having first provided some 
antidote to serpent poison in the form of some 
mantra or medicine. 

[The eating of the sacrificial cake by letting the teeth 
pierce it is prohibited as follows —na dato gamayet , yad dato 
gamayet, sarpa enatn ghatuka bhavanti. 

324 . Even if, sometimes, the truth is told in 
connection with praise or blame, the object is always 
to teach action or abstention from action. 



Having thus shown that individual words and their 
meanings are unreal and that only the sentence and its 
meaning are real, the author wants to point out that the 
sentence is the source of the individual word. 

325. The meaning of all individual words has 
its souice or basis in the meaning of the sentence. 
If the meaning of any sentence is incomplete, it 
is on a level with the meanings of the individual 

[The Vrtti, echoing what is going to be said in v.419-420, 
compares the relation between the meanings of words and 
that of the sentence to the relation between the senses and the 
body as a whole: . . . tathapi indriyanamiva Sariropanibandhanartha- 
kriyd vakyoparibandhanaiva sarvapadarthanam arthavyavastha. ] 

326. If a single word (a noun) is pregnant with 
the idea of action then, that also, they say, is a sen¬ 
tence, devoid of a verb, because (the understanding 
of action) is seen to take place from a sentence 

[The translation is based on the Vrtti text of the verse the 
second half of which is : antarena kriyaiabdain vakyad eva hi 
darUnat. This second half is missing in R and RP gives it in 
the footnotes with the remark that it is found in the manu¬ 
script gha and numbers the first half together with the first 
half of the next verse, though not for the purpose of 
translation. Though the Vrtti on 326 is by no means clear 
because of several gaps in it, it certainly seems to take 
this half as the second half of 326. I have translated the verse 

326(a) So is a mere verb said to be a sentence 
if a definite means for the accomplishment of the 
action denoted by it is understood and its meaning 
thus becomes complete. 

II. 327-328 


[It is a pity that the Vrtti has gaps here also. It seems to 
give examples of bare verbs which we mentally complete by 
supplying the necessary means for the accomplishment of the 
action denoted by them. Parjanya and MdtariSvan are mentioned 
as substrata of the iakti ( karakaSakti , kartrSakti)? This 
verse is also numbered as 326, (a) being added to distinguish 
it from the previous one, in order to maintain uniformity 
of numbering with R and RP.] 

Here the Mimamsaka objects as follows_ 

3 2 7. It is a certain completeness of idea, sepa¬ 
rated by and depending upon the words that are not 
used, which through inference, appears to be the 
cause of our understanding the meaning (of the 
words not used). 

[ When we hear the bare noun vrkfah tree we understand a 
complete meaning from it such as: ‘the tree stands’ or‘the tree 
is seen’. It means that we supply the word tifthati or drsyate 
mentally. This is irutarthdpatti. The point of the Mimam¬ 
saka is that a single word cannot convey a complete meaning. 
We supply some word or other mentally to complete the sen¬ 
tence and the sentence-meaning. In other words, the under¬ 
standing of the complete meaning is separated from that of the 
incomplete meaning of the word actually used by the unused 

word coming to the mind. asminneva fabde viii$tarupe yd 

buddhir utpadyate taya vyavahitam buddhyantaram buddhau prdpta- 
sannidhanam tadarthapratipattinimittam bhavati. ] 

The objection is answered as follows. 

328. Whatever meaning is understood when¬ 
ever a certain word is u ttered, that meaning belongs 
to that word (and to that word only). There is no 
more correct definition of meaning than this. 

[The point of the grammarian is that no part of the meaning 
should be attributed to any word supplied mentally. In other 
words, he does not admit Srutdrthdpatti .] 


VAKYAPAD1YAM of bhartrhari 

[The Vrtti begins its remarks on this verse with the words:— 
anyc tu many ante . It obviously means somebody other than the 
Mlmamsaka. It can very well be the grammarian and the 
author himself is one. His remark neha kaScit Sabdartha- 
sambandhasya kartd vidyate also confirms that the grammarian 
is meant. ] 

That being so the sutra :—kriyarthopapadasya ca karmani 
sthaninah (P. 2.3.14) is unnecessary, but it is based on srutar- 
thapatti= supplying a word mentally. 

329. That being so, in cases where a vei'b deno¬ 
ting an action meant for something else is the proxi¬ 
mate word, the action denoted by the unused verb 
is understood and so does the mere preposition like 
nih express the meaning of kranta etc. in a complex 

[The purpose of this verse is to show that the supplying of 
non-existent word mentally is unnecessary. In the sentence 
edhebhyo vrajati, the verb expresses an action meant for some 
other action, the word for which is not mentioned but is under¬ 
stood from the word edhebhyah = ‘for fuel’, that is, for bringing 
fuel. The action of bringing is understood from the word for 
fuel itself. That is why P.2.3.14 has been declared to be un¬ 
necessary, as the fourth case ending in edhebhyah can come by 
tadarthye catuarthi. Similarly, in nifkauSambih, nih by itself can 
express the idea of krantah. 

The Vrtti remarks as follows:— edhebhyo vrajatityasmin 
vakye edhebhya ityevamadina gatarthatvad aharanakriyartham prthak 
Sab do na prayujyate. ] 

If half a sentence can express the meaning of the full 
sentence and both are sentences, what is the use of ever using 
the full sentence ? 

330. They are separate expressions and are like 
synonyms. The meaning of sentences consisting of 
single words is determined by purpose and the 

II. 331-333 


[ When the sentence consists of a single word only, noun or 
verb, how to complete it mentally would depend upon pur¬ 
pose and context. 

The Vrtti points out that Vrk$ah and Vrk$as tifthati are 
both complete sentences and not that the former is the latter 
devoid of the word ti$thati. Both convey a meaning. They 
are like the pair Vrsa and Vrsabha, or Tdva and Tdvaka or 
praparna and prapatitaparna in which each of the two is com¬ 
plete and not that the former is an incomplete form of the 
latter:— tani tvaprayujyamdnapadaikcideSasarupdni samuddyd - 
ntarani vrsavr$abha—ydvaydvaka—praparna prapatitaparnavad \. 

If only the sentence is real, what is the use of analysing it 
into words ? 

331. The means whereby the meaning is under¬ 
stood and which differ in the case of each person are 
not in any way connected with the thing to be 
known from the sentence. 

[ The idea is that the artificial divisions are the means of 
understanding the integrated sentence. 

332. Even when no meaning is understood or 
when a wrong meaning is understood words are still 
eternally connected with their meaning. 

[The relation between word and meaning is eternal. That 
is not affected by the fact that the listener may not under¬ 
stand the meaning due to absent-mindedness or due to igno¬ 
rance. ] 

333. When the word dvaram (—the door) having 
the second case-ending is heard, then, according 
to the context one understands either ‘shut’ ( badhdna ) 
or ‘give way' ( dehi ). 

[This is said in support of srutarthapatti. According to the 
context, one is entitled to supply mentally some word or other 


VAKYAPADlYAM of bhartrhari 

in order to complete the sentence. Otherwise, one would not 
understand meaning and the relation between the word and the 
meaning would be affected. ] 

The opponent points out a disadvantage in not accepting 

334. A word expressive of a finished thing 
( sattva ) and standing for the means (whereby an 
action is accomplished) cannot also express the main 
meaning of a sentence, namely, the action to be 

[The point sought to be made is that to avoid the disad¬ 
vantage, the required word must be mentally supplied. In 
other words, srutarthapatti must be resorted to. ] 

335. Therefore, when a sentence consists of a 
single word (whether that word is a noun or a verb) 
it does no more than express its own meaning and 
retires incomplete. Its meaning suggests the proxi¬ 
mity (of the other meaning required for complete¬ 
ness through the medium of the word to which the 
meaning belongs.) 

[In this way, Srutarthapatti is justified by the Mimamsaka. 

The upholder of indivisibility now criticizes the Mimairisaka. 

336. As there is no difference in the matter of 
being meant for something else, the word (actually 
uttered) cannot bring another word to the mind, 
nor can its meaning bring it nor can the uttered 
word bring the meaning ( of the non-uttered word) 
to the mind. 

[The word actually uttered cannot bring the word not uttered 
to the mind, because its function is to convey its own meaning. 

II. 337-338 


Nor can its meaning bring the unuttered word to the mind, 
because there is no relation between the two. Nor can the 
uttered word bring the meaning of the non-uttered word to 
the mind because there is no relation between a word and the 
meaning of another word. A fourth possibility is that the 
meaning of the uttered word brings the meaning of the non- 
uttered word to the mind, but that would be a case of infer¬ 
ence. The conclusion is that when a word, noun or verb, is 
uttered, it brings to the mind without the intervention of any 
other word but with the help of the context another meaning 
which is required to complete it. In this way, the Mlmam- 
saka view is rejected. 

The Mimamsaka replies:— 

33 7. If the word expressive of the object (of an 
action, dvdram in the present case) makes the verb 
non-existent as it were (nasiarupam iva ) and (there¬ 
fore) unnecessary, then both action ( bliava ) and the 
finished thing ( sattva ) would acquire equal impor¬ 

[The Grammarian himself believes that both action and the 
finished thing cannot be equally important in a sentence. If 
the actually uttered word dvdram conveys the meaning of the 
verb also which is not used, then both action and the finished 
object would be conveyed by it on an equal footing which is 


Now the Grammarian says:— 

338 (ab) They describe it (the word dvdram ) as 
a verb similar in form to a noun. 

If that is so, how to account for the second case-ending in 
the word ? 

(cd) Usage is regulated by the positive and the 
negative reasoning (anvaya and vyatireka ). 

[The word dvdram is really a verb similar in form to a 
noun. Like other verbs, therefore, it also denotes action 



primarily. If one sees the second case-ending in it, it is only by 
analogy. In sentences lik ebhandam pidhehi , dvaram pidhehi where 
both noun and verb are present, one sees that the verb denotes 
action primarily and the noun denotes substance and takes a 
case-ending. Where only the noun is used but both action and 
substance are understood, the word takes a case-ending be¬ 
cause it denotes substance also. 

Here the Mimamsaka asks a question : If the word dvaram 
denotes more than one thing, is not its expressive power there¬ 
by lost as there might be a doubt as to which meaning to take ? 

The grammarian answers— 

339. Even though doubt arises on account of 
(similarity) of form, still its expressive power does 
not disappear as in the case of the expression ardham 
pasoh (=half of the sacrificial animal) where compe¬ 
tence decides what the meaning is. 

[In ardham pasor iva, because of the word iva, a doubt 
arises as to whose half is meant, of the animal or of Devadatta, 
but the doubt is resolved with the help of context and appro¬ 
priateness. Similarly, where only the word dvaram is used, 
context and appropriateness would help one to decide whether 
it should be taken as a noun or as a verb.] 

Mimamsaka objection:— 

340. If, when a word expressive of sattva (con¬ 
crete object) stands by itself it is to be considered as 
being expressive of a process ( bhava ), if, when they 
come together the word expressing sattva is consider¬ 
ed to have a different meaning and if the verb does 
not have its own separate meaning, 

341. Then, the statement (of old thinkers like 
Yaska) that a verb expresses chiefly a process whereas 
a noun denotes a concrete thing ( sattva ) and that 
there are four parts of speech would be contra¬ 

II. 342-344 


[There is a reference here to Yaska’s definition in his 
Jsfirnkta of the verb and the noun : bhavapradhanam dkhydtam, 
sattva-pradhanani namani. It is also in the Nirukta that one 
finds the earliest mention of the four parts of speech : ndma- 
khyatopasarganipdtah. See Yaska Nirukta I. 1.] 

The Grammarian answers— 

342. Having seen the eternality of the sentence 
in the mind and its connection in the world with its 
meaning, Varttaksa and Audumbarayana have 
declared that there are no four parts of speech. 

[When the hearer grasps the indivisible sentence, he is 
not conscious of any parts in the form of individual words. 
Nor is he conscious of word-meanings when he grasps the sen¬ 
tence meaning consisting of pratibhd. 

As the sentence is indivisible the individual word does 
not exist and so the question of the different kinds of words or 
arts of speech does not arise. Those who believe in srutdrtha- 
patti would naturally accept divisions within the sentence and 
their being brought under four classes., 

What then is the function of the Sastra ? 

343. Both in the world and in the Science of 
Grammar, the easy and very comprehensive treat¬ 
ment of sentences on the basis of individual words 
is resorted to by convention for the sake of conve¬ 

The Mimamsaka objects—If the individual words and 
meanings are non-existent, how is it that action and abstention 
from action based on them take place in the world ? The 
Grammarian answers— 

344 . In the world, traditions of men are not 
always based on reality. Therefore, there is nothing 
beyond the sentence which is not a mere creation of 
the world. 



[The point is that when we are trying to understand the 
truth, we cannot rely on the practice of the world. In the 
world, one sometimes reaches the real through the unreal. We 
can understand what kind of animal a cow is by looking at a 
picture of it. That does not mean that the picture of a cow is 
a real cow. Similarly, we sometimes act on the basis of word- 
meanings but that does not mean that they are real. ] 

345. Even when the meaning of a sentence is 
made clear by means of indications, found in other 
sentences, that meaning is its own. It was not clearly 
noticed on account of identity of forms. 

[The Sruti says : aktah sarkara upadadhati= he places 
sugar mixed with grease. 

It is not clear from this sentence whether the sugar is to 
be mixed with oil or clarified butter. But another Sruti says : 
tejo vai ghrtam= ‘clarified butter is indeed lustre’. That is a 
clue, an indication that the sugar is to be mixed with clarified 
butter. The first fruti thus means : ‘he places sugar mixed 
with clarified butter’ and this meaning belongs entirely to the 
first srati in spite of the fact that the clue is found in the second 

Something is now being said about the relation of badhya 
and badhaka (what is set aside and what sets aside) between 
the original rules and their exceptions. 

346. Whatever is excluded implicitly, if not 
explicitly, from the operation of the general rules, is 
explicitly mentioned in the special rules of exception 
and that is really part of the meaning of the general 

[A general rule does not really cover special cases though 
there is nothing in the rule itself to tell us so. But when the 
author of the general rule framed it, he knew about it. 

An example of the same from the world is now given— 

347. The injunction regarding the serving of 

II. 348 


curds to Brahmins operates by excluding Mathara. 
The word Mathara (in the subsequent sentence 
takram mdiharaya) because of its connection with 
butter-milk, only confirms the correctness of the 
meaning of the previous sentence (brahmanebhyo 
dadhi dlyatam). 

[What is pointed out here is that the sentence takram 
mathardya does not set aside the giving of curds to Mathara 
because that was never envisaged by the first sentence brah - 
manebhyo dadhi diyatdm which enjoins the giving of curds to 
Brahmins excluding Mathara. So there is no question of 
contradicting something which was enjoined before. Serving 
of curds to Mathara was never enjoined. 

Thus what is called bddha of the first sentence by the 
second sentence is nothing more than the inferential under¬ 
standing that curds were never meant to be served to Mathara 
aprdptyanumdna . 

What was said in the previous verse was based on look¬ 
ing upon the two as two separate sentences. It is thus that 
bddha amounts to aprdptyanumdna : The second sentence be¬ 
comes the cause of our inferring that the injunction in the 
first sentence was never meant to apply to Mathara. 

If the two sentences are treated as one what would be 
the nature of bdidhyabadhakabhava ? 

348. Even though there is more than one verb, 
some look upon the original injunction and the ex¬ 
ception to constitute one sentence which appears to 
be divided. 

[If they are looked upon as one sentence, the exclusion, 
of the sphere of the exception and the application of the 
original injunction to the rest would take place at the same 
time and so there would be no need for aprdptyanumdna. 

Just as an exception is supplementary to the original 
injunction, in the same way, restrictions and prohibitions can 
be looked upon as supplementary to original injunctions. 


vakyapadIyam of bhartrharj 

349. That being so 3 restrictions and prohibitions 
are also supplementary to the original injunction. 
The aluk (which is taught in P. 6.3.1) has been 
declared to be supplementary to the luk (which is 
taught in P. 2.4.71). 

[Exceptions, restrictions and prohibitions are looked 
upon as making one sentence with the original injunction on 
the basis of either vakyaikavakyata or padaikavakyatd. The former 
is defined thus :— upakaryopakarakabhavapannabodhajanakatvcm . 

Without using these technical expressions, the Vrtti ex¬ 
plains the same ideas in its own words : vidhikdla evotkrftasya 
punahSruter apraptir anumiyate = At the time of the original injunc¬ 
tion itself, it is inferred that it would not apply to what 
would not come within its scope. Similarly when P. 2.4.71. 
teaches the elision of the case endings of words which are to 
enter into a samdsa, it is understood that this elision does not take 
place when particular words become the second term in a com¬ 
pound. In other words, the original injunction and the later 
exception or restriction or prohibition are understood as one 
sentence and so understood together : yasca supo dhdtuprdti - 
padikayor iti dvitiyena lug anvdkhydyate tasya tasminnevavadhivakye 
uttarapada-vi§e§am varjayitveti prathamameva vidhina prakalpitah 

Those who are against looking upon them as one sentence 
argue that the sentences in question are complete in their 
meaning and so there is no relation of primary and secondary 
between them. 

350. The sentences in question are independent 
as they do not require one another when they con¬ 
vey their meaning. As they do not serve one an¬ 
other, how can they constitute one sentence ? 

[The Vrtti points out that two separate sentences can make 
up one sentence only if there is some kind of dependence or 
mutual requirement between them. If each is self-sufficient 
it is concerned with itself only. If they cannot render service 

II. 351 


to each other, there is no question of their making up one 
sentence:— Ilia sdkar.ksanamsamsargat parasparam upakare vartama- 
ndndmekavdkyatvam upapadyate. Pradhandni iu prthagdtmanirvrttau 
vyaprtdni. Tesdm nirakanksatvad asatyupakdre ndstyekavdkyatvam.] 

The upholder of the single sentence argues as follows— 

35 1. (As the original injunction) requires the 
special injunction, the latter is taken to be supple¬ 
mentary to it. The special injunction also requires 

the other in the matter of determining what is ex¬ 
cluded by it. Therefore, mutual requirement is 
similar in each case. 

[What is pointed out here is that the original and special 
.'unctions ( n iyama, apavada or pratisedha) require each other 
'• order to determine their own meaning. As their mutual 
requirement is equal, it is better to look upon them as consti¬ 
tuting one sentence. Each requires the other in order to ex- 
clude & the scope of the other from its own scope. 

While explaining this verse the Vrtti takes a different kind 
of example altogether. It takes the sfitra P. 3.1.96, that is, 
tavyat-tavya-aniyarah = ‘After a root occur the suffixes tavyat, 
tavya and aniyar. It says that according to some, the verb ‘to 
occur’ is one though it has three different agents. The verb 
° . one, there is only one sentence here :— bhavater ekatvad 
ekavdkyatvam. According to others, the action denoted by the 
verb becomes different with each different agent. With each 
agent, the action is self-sufficient and so there are as many act¬ 
ions as there are agents and so as many sentences. Those who 
uphold the view' that there is only one sentence say that each 
of the three suffixes is separately the agent of the verb bhavati, 
while mutual requirement does exist. The agent-power (kartr- 
jakti) is one, but it exists in three different suffixes : ekavd- 
kyavddinastu manyante—satydm apeksdydm tavyadddayah prthak 
prthag bhavateh kartdro vijndyante. Bhinnddhdrd vd tavyadddindme- 
kaiva kartrlaktir iti.~\ 

The question whether individual w'ords can also be analysed 
into parts is now' discussed:— 


vakyapadIyam of bhartrhari 

352. There is no elision of the part of a name. 
A name with a particular sequence of phonemes has 
been given to something and that does not disappear. 

[Devadatta is a name. Some people sav only Deva or 
only Datta. That may happen in the world but the lustra does 
not teach the elision of either Deva or Datta. As the Vrtti 
puts it : ca visistarupaydin krtdydrn sovijiidydvi nirjiidta- 
pravrtti prasiddhaprayogam rudharupam punar anyalhd sakyam 
kartu/n .] 

353. From Delta etc, a different name alto¬ 
gether, another name (Devadatta) cannot be under¬ 
stood. Nor can Datta bring to the mind the named 
(individual) Devadatta, because it is the named of 
some other name. 

[A part of a name can neither bring the full name to the 
mind nor bring the named of the full name to the mind. The 
Vrtti says the same thing in its own way. Ilia yo’rthena krta- 
sambandhah Sabdah sabdantaram tasya vdcako na bhavati = When 
there is a relation between an object and a word, another word 
cannot be expressive of that word, that is, another word can¬ 
not bring that word to the mind. All this has been said in 
answer to the suggestion, that Datta first brings Devadatta to 
the mind and, from the latter, the individual Devadatta is 
understood. The possibility of this process is denied here.] 

Others believe that the ‘named’ is connected not only with 
the full name but also with parts or abridgements of it. 

3 54. Others are of the view that there is a rela¬ 
tion between the named and all the parts of the 
name as with the full name. 

[ In other words, a part of the name can bring the ‘named’ 
to the mind as well as the full name can. Or rather, the whole 
and the parts bring the 'named 5 to the mind at the same time. 

As the Vrtti says .... iti tatra sanijiidkale sarva eva sarnjnino 
yathaiva scimudayais tathdvayavair api smnbadhyante = at the time 

II. 355-357 


that the name is given, the named individuals are connected 
with the full name and parts of that name at the same time.] 

Two defects are pointed out in this view:— 

35 5. In this way, even the phonemes which are 
parts of the name would become expressive of the 
named. Nor does one see the expressive power of 
the part once it is separated from the whole. 

[The two defects are : (i) If a part of a name can denote 

the ‘named’, why not each phoneme of it. ? (2) Whatever ex¬ 
pressive power a part has exists only as long as it is part of the 
whole, not when it is separated from the whole :— sambaddhefu 
vavayavcsu samjndtvam pratipadyamane'u kevald daitddaya uccdrya- 
manah samjiidtvcin na pratipadyeran dvirvacanavad eva iti, says the 

356. If the whole name, together with its parts, 
is expressive of the ‘named’, then it is not possible 
for what looks like parts to be expressive of it. 

[What is emphasised here is that what looks like a part 
is really not a part. It only looks like it. 

The Vrtti points out that the parts which depend upon the 
whole cannot perform the same function apart from the whole 

._ na hi samuddyatantramm avayavdndm ekdrthakriydydrri hitvd pra- 

dhdnam pravrttih sambhavati .] 

Another view is now given. 

35 7. According to some, from the part the 
whole is remembered. From the whole thus remem¬ 
bered, the meaning of the whole is understood. 

[The Vrtti points out that it is not merely a word that is act¬ 
ually heard which can convey a meaning. A word which is 
regularly inferred can also, like the word which is heard, con¬ 
vey a meaning :— Kecit tu manyante—navaSyavi iruyamana eva 
sabdah pratyayakah. niyamendnumiyamdno 5 pi sruyamanavad eva 
pratyayam ntpddayati. From the parts, perceived apart from the 


VAKYAPADlYAM of bhartrhari 

whole, one does remember the whole and the whole thus re¬ 
membered, does cause the cognition of the object named :— 
tatha ca smaryamanah sa eva krtasambandhah samuddyah samjninam 
pratyayayati . 

That view is now refuted. 

358. How can there be remembrance from the 
part of the whole which is different from it ? How 
can a word which is only understood (and not 
heard) be expressive of its meanings; 

[The whole is different from the part. So, on seeing the 
part, one cannot remember the whole. Even if one does re¬ 
member it, it cannot convey the meaning because the whole 
has not been uttered or heard. To believe that what has not 
been heard can convey a meaning would result in undesirable 

The Vrtti points out that what looks like a part can be a 
part of many wholes having different meanings. Which whole 
would one remember on seeing the part ? :— sddhdrano'asdv 
ekadesas tulyena rupend nekasamghdidnupatx . Tatra bhinne$u sam- 
ghatesu niyamat katham smrtih ?] 

Now the siddhanta is set forth. 

359. (When names are applied to objects) 
words similar to their parts and endowed with their 
different characteristics arise later and are applied 
as names to the same object. 

[The Vrtti points out that when full names are given to ob¬ 
jects, other names, similar in form to parts of the full names, 
come into existence as by-products. The process is similar to the 
coming into existence of by-products when we try to obtain 
any main product; Ke$ancittu samudayasvarupe samjiidtvena prakal- 
pyamane samjndntaranyeva parikalpitaikadesarupani ekasadhandn - 
yanuni§padyante . Niyata ceyam apauru$eyikasarvasamjndsamjiiini 
sambandh avis ay a sab dan tarandmek adesasar up an dm an u nispatt i r i ti . 

What is pointed out is that these words which arise later 
(anunispattih) are not really parts of the name, but only look 

II. 360-361 


like them. According to the doctrine of indivisibility, they 
cannot be parts.] 

These words which arise later and look like parts of names 
sometimes cannot invariably convey the individuals expressed 
by the names themselves. Deva which looks like apart ofDeva- 
datta can mean a god or the person whose name is Devadatta: 
The form is the same in both cases and so a doubt can arise. 
How the doubt is resolved is now stated— 

360. The form being common, their meaning is 
doubtful and their substratum (in the form of mean¬ 
ing) is determined through competence ( samarthyat). 
The sastra teaches elision etc. in the case of those 
which are grammatically correct. 

rof the words which arise later, some like Deva and Datta 
are grammatically correct and the Sastra teaches elision (lopa) 
etc in their case. Others likej>, dr a, kha are grammatically 
incorrect. It is in the case of the former that the meaning can 
be doubtful because of similarity of form. The doubt is re¬ 
solved through context taken in its widest possible meaning. 
The sastra also says something about their derivation. But not 
about the derivation of jye, dra, gha etc. which are therefore 
'ected. As the Vritti puts it— tatra sadhavo lopasdstrendnvydkhya- 
X yam anarupaparicchedah parigrhyante. Tat o' nye tyajyanie. ] 

361. Though there is similarity as far as arising 
later is concerned, dra, kha are incorrect. There¬ 
fore, in the sastra which explains derivation, they 
are not mentioned as Datta etc. are. 

[According to the Vrtti, the incorrect forms are jye, dra and 
kha abbreviations of jyeftha, ardrd and visdkha, all names of 
stars. But these abbreviations are mere collections of 
phonemes and not words and, therefore, meaningless. There¬ 
fore, the sastra takes no notice of them: Tatra ye ndnvdkhyatdh 
sastrend tulyaydmanu nispattau jyefihdrdrd visdkha ityevamadisu 
jye-dra-kha ityasadhavo vijhayante. JVa hi te ’nvakhydyake smrtisastre 



dattadivat smaryante . There were apparently some who be¬ 
lieved that once the whole is taught as a name, its parts are 
also automatically taught as names. In other words, it is not 
necessary to derive the parts separately as names:— Tathdsyapi 
tantrena prasaiigena vdsamudayasya samjiidtvenopaddne tadavayavdna- 
mapi samjndtvam vyavatisthate.] 

A doubt as to the eternality of the relation between names 
and the named is removed as fellows— 

362. Names like kharanasa have n instead of n 
eternally and as they denote a particular individual, 
they have been declared to be correct. 

[The woid kharanasa has this form when it is not a name 
but it becomes kharanasa when it is a name. They are really 
two different words. It is not that the word kharanasa be¬ 
comes kharanasa when it is to be used as a name. So the 
question of the non-eternality of the relation between the 
name and the named does not arise at all. 

The Vrtti quotes a verse stating the doubt to resolve which 
the present verse has been given. The doubt is: If the word 
kharanasa is formed before it was a name then how did it get n 
which depends upon its being a name ? If the word was khara¬ 
nasa when it was formed and it was changed to kharanasa when 
it became a name, the eternality of the relation is 
affected. To resolve the doubt, it is stated that forms 
with na and forms with na are both eternal. One is 
not a transformation of the other:— krtanatvdscdkrtanatvdka 
nityah samudayd vidyante. Tatra krtanatvah samuddye niywiyamd- 
nah sddhavah. Anyatrakrtanatvdh sddhavah. Whether with n or 

with ? 2 , they are both incorrect if used outside their scope:_ 

nbhayesam ca visayaviparyaye sadhntvam na vidyate. ] 

3 63. These words (like kharanasa) are names of 
individuals because the power to be a name inheres 
in them. In such cases one does not necessarily 
look for the presence of the corresponding charac¬ 
teristic (nimitta ) in the named. 

II. 364-365 


[What is meant here is that even though the word kharcinasa 
means ‘one who has a sharp nose 5 , one does not look for a 
sharp nose in a person before giving him that name. It can 
be applied to any body. The power to be a name exists in 
that word because of the n in it. A word having a particular 
form is given as a name to an object irrespective of whether 
that object has any feature corresponding to the meaning of 
the name: Kdnicit tu nimittasydbhave viparyayena svariipamatra - 
nibandhanani sannipatanti says the Vrtti .] 

The fact that names are given to objects by persons accord¬ 
ing to their wish does not mean that the relation between the 
names and the named is not eternal. 

364. For the sake of worldly transactions some 
names are restricted to particular objects in parti¬ 
cular places. But the relation in the case of a name 
like JDittha is as eternal as in the case of the name 


[Even though we see persons giving names to objects, it is 
not they who create the relation between these names and 
the objects. That relation, being eternal, was already there. 
Persons only make use of it in particular places and times. 

The Vrtti points out that all names have the capacity 
to denote all objects and all objects can be called by all the 
names : samjndndm sarvasamjnipratydyanavi$aydh iaktayo vidyante. 
Sarnjnindmapi pralyekam sarvasamj ndpadavisayatdkhyah saktciyo 
vidyante. It is only for the sake of convenience that particular 
names are given to particular objects in particular places. 

In the Sdstra also, names are restricted to particular things 
in order to facilitate grammatical operations. 

365. In this Sastra , the non-artificial relation of 
names like Vrddhi (with the named) consists in res¬ 
tricting their power to particular objects. This 
relation is like the relation of the qualifier and the 

[All words are capable of being the names of all things and 



all things are capable of being named by any word. But in 
the Sastra , a particular name is restricted to a particular thing. 
In other words the relation of the name and the named is 
eternal. All that human beings can do is to restrict it in 
particular cases for the sake of convenience. The person who 
says blue jar does not create the relation between the blue 
and the jar. The expression nilo ghatah only states a relation 
which was already there: tad yathd nilam iti sarvaSrayaviSe$ana - 
Saktir utpaladivisaya paiigrhyate. Utpaladinapi sarvaviSe$anavi$ayd 
viSe$yaSaktir niladi$uniyamyate= when one says nilam utpalam, the 
power of blue to qualify any subtance is presented as relating 
to the lotus and the power of the lotus to be qualified by any 
quality is here restricted to the blue, says the Vrtti. 

Something is now said about the two kinds of names. 

366. A name is given on the basis of its own 
form, sometimes when there is some circumstance 
in the object corresponding to the form and some¬ 
times when there is not. 

[Both in the world and in the Sastra names are given mainly 
on the basis of their form. Sometimes, a name is said to be 
anvartha when there is some circumstance in the object corres¬ 
ponding to the form of the name as when a person who is 
good at destroying his enemy is called Satrughna. But there is 
no need for it: for example, the name Dittha in the world and 
the names ti, ghu , bhd etc. in the Sastra . 

Not only small names consisting of just one syllable but 
also big names are given in the Sastra on the basis of form 

367. In th c sastra, a big name is based on its 
form but when some circumstance is present in the 
object, that is also inferred to be the basis of the 

[When the form only is the basis of the name, the mean¬ 
ing of its parts is not considered to be the basis. But when 
there is some circumstance in the object and it is expressed by 

II. 368-370 


the parts of the name, then the meaning of the parts also be¬ 
comes the basis of the name.] 

How it is inferred is now stated. 

368. Because of repetition of form, its repeti¬ 
tion is inferred. Or it is understood as another word 
altogether or a difference of power is understood. 

[In the case of the long names found in the Sastra , three 
possibilities are envisaged: (1) it should be taken twice, 
once as a whole having a particular form and again as a whole 
made up of parts through the meaning of which it denotes 
the object, (2) it should be taken as two separate words, 
(3) it should be taken as one name having two distinct 
powers. Both the Vrtti and Punyaraja speak about these three 
possibilities, the former in a somewhat obscure language and 
the latter a little more clearly.] 

It is now shown that there are four possibilities in the 
matter of names in the Sasira. 

369. Sometimes, a technical name is applied 
differently in different contexts. In the case of the 
name Samkhya, both the technical and non-techni- 

meanings may be applicable in the same text. 

[In P. 1.3.14, the word karma has a non-technical meaning, 
namely, action, but in P. 3.2.1, it has its technical meaning 
namely karmakaraka = the object of an action. In P.5.1.22, 
the word samkhya has to be understood in both its technical 
and non-technical meanings.] 

3 70. Sometimes, a worldly name when used, 
covers the field of the technical name also, as the 
word sambuddhi in P. 1 . 2. 33, where it should be 
taken both in the technical and non-technical senses. 

[,Sambuddhi , in the world, means calling or addressing some¬ 
body. Technically, it is the vocative singular by P. 2.3.49. In 
P. 1.2.33, both the meanings apply according to this verse. 


VAKYAPAD1YAM of bhartrhari 

Compare, however, the following remark of the kafikd on 
P. 1.2.33— Durat sambodhayati yena vakyena tat sambodhanam 
sambuddhih. Naikavacanam sambuddhili . The kdsikd seems 
to exclude the technical meaning here but it does not really 
do so. All that it means is that while understanding 
the sutra> the non-technical meaning should be taken and 
not the technical one, though the latter is also included in it. 
The Vrtti with a little correction in the text, makes this point 
clear— Diirdt sambuddhav ityatra laukika eva kriydSabdali krtrimdyd 
api vi$ayam sannidhanad vydpnoti. If the sutra is understood by 
taking the technical meaning of the word, it would not cover 
all cases:— krtrimd tu na saknoti lydptum sarvakriydh Sabda- 

It is the sentence which is being discussed in this part of 
the text. One point which is discussed is whether the action 
conveyed by the sentence is carried out individually or by the 

371. Scholars declare that when expressions 
relating to the class are used or when there is reten¬ 
tion of one or when the dvandva compound is used, 
the actions are related to each one of their substrata 
because of the nature of the word itself. 

[By substrata ( dsraya ) of the action, the agents are meant. 
The examples of the three cases envisaged are : (1) brdhmand 
bhojyantdm (samgha = class). (2) brahmano bhojyatdm {eka$e$a), 
and (3) Devadatta-Yajnadatta-Vi$numitrd bhojyantdm {dvandva). 
In all these sentences, the action of feeding has to be applied 
to each agent separately.] 

It is now stated why an action like eating cannot rest in 
many at the same time. 

372. The action of eating, consisting of a parti¬ 
cular form and a particular result, rests in each 
person separately. Otherwise the meaning of the 
root bhuj =to eat, would not be carried out. 

[The action of eating is not like the action of dramatic 

II. 373-375 


representation ( natya) which is a name applied to the different 
actions of many taking place at the same time. The name 
‘eating’ is applied to a particular action of a person, resulting 
in his satisfaction ( trpti) . If its parts and its result are found 
in different individuals, it would not be called eating at all. 
As the Vrtti puts it— Tadi hi bhinncsu bhoktryu anyatra rupain 
rupaikadebo vanyatra drSyate, anyatra phalatn phalaikadcso va vyava- 
tistheta bhujir evasau na syat.~\ 

3 7 3. All eaters individually perform the act of 
taking the food and so on and attain the result 
namely satisfaction and not (collectively) as in the 
case of dramatic representation. 

[The different subsidiary acts which constitute a dramatic 
performance are meant to be performed by different indivi¬ 
duals and it is only when they do so that it is said to be well- 
done : As the Vrtti puts it— Tathaiva ca sa nirvartyamana samya- 
ganufthitetyucyate. This is not true of eating where each indivi¬ 
dual has to do all the acts like taking the food, moving the 
jaws etc. himself, right till final satisfaction: Annadanahanucala- 
nariipo hi bhujis trptiphalo bhoktrbhih pravibhaktali parisamaptaru- 
paphalo nirvartyate .] 

Another illustration is now given. 

3 74. It (the action of eating) is like the water 
for washing the feet of guests; by its very nature 
( samar thy •at) it is arranged separately for each person. 
The action of eating does not produce its effect 
(satisfaction) for several at the same time as a lamp 

The difference between the action of eating and the action 
of seeing is now pointed out. 

3 75. An action like seeing is one and without 
being repeated in each case it can cover all its ob¬ 
jects which are suitably situated at the same time. 


VAKYAPAD1YAM of bhartrhari 

[An action like eating can cover many agents only if each 
agent performs that action separately. But the action of 
seeing can cover many objects at the same time, provided they 
are situated in the right place. Of course each object can be seen 
separately also but the point here is that they can be seen at 
the same time also. As the Vrtti puts it— Vanam drsyatom, 
sangho dr Sy at am iti drsirupasyobhayatha darSanasambhavad ekam vd 
darsanam samudayakarmakam nirvartyate, pratyavayavam vd kramena 
bhinnadarsanamavartate. ] 

The question now arises as to whether the name ‘cooking’ is 
applied to one particular act or to a series of acts understood 
as a whole. 

376. The perception of the different accessories 
( karaka ) witli their different activities when one 
observes worldly transactions makes one conclude 
that the action (of cooking) covers the activities of 
the different accessories taken together. 

[Two views are mentioned about the action of cooking. 
One view is that it stands for the softening of the material that 
is cooked, that is, for viklitti. Another view is that it stands for 
all the activities taken together of the different accessories 
which play apart in the cooking, such as the cook, the fuel, the 
pot and so on.] 

It is now shown that in the Sastra also sometimes the 
sentence-meaning is applied to each case separately. 

3 7 7. As the form to be derived is well-known 
in the world and as there are indications in the 
sastra itself (for separate application) and as each 
item requires application, one applies the name 
Vrddhi to a , ai and au separately. 

It is now shown that sometimes the meaning of the 
sentence is applied to the group as a whole. 

378. As the taking of one hundred is the main 
thing in an order relating to the realisation of a fine 

II. 379-381 


of hundred, even though the persons to be fined are 
many the amount of fine to be realised by the au¬ 
thorities does not vary. 

[According to the order : Gargah Satain dandyantam = Let 
the Gargas be fined a hundred, all of them are collectively 
fined hundred and not each one hundred separately. Because 
the total amount to be realised is hundred and that is the 
main point in the order. If each one is fined a hundred the 
total would exceed hundred and the order would be violated. 
As the Vrtti puts it— Tatradanasyapi Satdrthatvdt tadarthyena 
$ a t am pradhanam na gunabhedad bhidyate . . . Yathaiva Sat am 
dandyantam ityukte na sahasram dandyante tathd pratyekam satam na 
dandyante. ] 

It is now stated that in the Sastra also, the meaning of the 
sentence is applied to the whole. 

3 79. As, in order to teach some grammatical 
operation, it is the name of the whole that is taught, 
the names samdsa and abhyasta are connected with 
them (the wholes ) by tantra. 

[T antra means the power of a word, uttered only once, to 
convey more than one thing. The siltras : saha-supa (P. 2.1.4.) 
and ubhe abhyastam (P.6.1.5.) teach the names samdsa and 
.abhyasta respectively. The words saha and ubhe show that the 
names are to be applied to all the items together.] 

It is now stated that sometimes the meaning of a sen¬ 
tence is applied to the whole as well as the parts. 

380. When a word standing for a certain 
property denotes many things in relation to a certain 
action, it does so severally and collectively. 

381. According to the order : Sudras should not 
enter this house, their entrance is prohibited seve¬ 
rally and collectively. 

164 vakyapadIyam of bhartrhari 

The same idea is now expressed differently. 

382. When a collective prohibition in regard 
to earning and the like is expressed, merely because 
the prohibition is not stated severally, its application 
severally is not contradictory. 

[What is pointed out is that the use of the singular or 
the plural number in prohibitory orders is accidental. In 
either case it can apply severally or collectively. 

The Vrtti points out that when the prohibition is collec¬ 
tive, it is the collection or the group which is the object of the 
prohibition. As to the individuals, it is left to them to do 
what they like. . . Samhananatmadharmasya irutyd prakrantatvat 
sanghata eva lipsatikriyasadhanabhavena pratisiddhyate. Pratyekam 
tu kamacarah .] 

An illustration from the sastra is now given. 

383. In the same way, as in P. 8.4.2. at, ku, pu, 
ah and num are causes of separation (vyavaya) whether 
they do so severally or collectively, they do not 
prevent the change of n to n. 

[P. 8.4.2 does not say in so many words that at etc. do 
not prevent the change of n into n when they come between 
the nimitta and the nimitti individually or collectively. The 
fact is that either way they do not prevent the change. As 
the Vrtti puts it— Tatha ca pratyekam samastair vd vyavaye natvairi 
bhavatyeva .] 

Whether the action of eating is one or many is now 

384. When the act of feeding is done for favour¬ 
ing those who eat, then, if there is no difference in 
time, place, etc. it does not favour them. 

[The idea seems to be that it is not by feeding them 
irrespective of time and place that one favours them. One would 

II. 385-387 


favour them either way, by observing difference of time and 
place and by not observing it :— Bhedena tu pratyekam prakrame 
sarvo’ nugrhy ate, says the Vrtti. If the feeding is for favouring 
the eaters, one has to act in such a manner that they are favour¬ 
ed, by not observing difference or by observing difference, 
according to what is needed : Tathd tesdm anugraho bhavati tatha 

viniyogavyapeksa. ] 

After this preliminary remark, the question is considered. 

385. The action of eating which is one has been 
declared to be many because of difference in vessels 
etc. Or the action of eating which is really varied 
(because of difference in the satisfaction of the eat¬ 
ers) is thought of as one (because of identity of 
time, place etc). 

[Thus difference of opinion on this matter has been 

The final position is now stated. 

386. The eaters, mentioned separately and act¬ 
ing together, eat their food served separately but 
together (that is, at the same time). 

What has been said in v. 378 is now further explained. 

387. The fine of hundred is to be applied to 
the group (of Gargas) because (1) there has been 
no separate mention (of the persons to be fined), (2) 
otherwise the amount to be realised would conflict 
with the other amount (that is, the amount actually 
realised) (3 ), the meaning of thesentence cannot be 
applied in a two-fold manner. 

[Gar gall satam dandy antam is the order and not Gargyo 
gargyo satam dandyatdm . Gargya has not been repeated. Instead 
of that, the ekafefa has been used. If each Gargya is fined a 
hundred, the amount realised would far exceed hundred which 



is the amount mentioned in the order. Lastly, the order can¬ 
not be applied in two ways, individually as well as collective¬ 
ly. By two ways, Punyaraja understands that the verb cannot 
be connected with the main object (hundred) as well as the 
secondary object (the Gargas). The conclusion is stated 
by the Vrtti which, otherwise, is none too clear, as follows— 
Tasmat s&nghdtci Bvaikain §Q.tcan avo,ti$thcLte'=P\, hundred has to be 
realised from the group (of Gargas).] 

Sometimes, however, the meaning of a sentence is connect¬ 
ed both individually and collectively. 

388. Where the action of eating is ordered to 
be performed together with others, either through 
the dvandva compound or through ekasesa, there also 
the meaning of the sentence is to be connected in a 
two-fold manner because the verb conveys its mean¬ 
ing through laksana. 

[Sentences like the following are here kept in mind I— 
Devadatta-Yajnadatta-Vismmitraih saha bhoktavyam=T)evadatta, 
Yajnadatta and Visnumitra must eat together. Here the action 
is ordered through dvandva. In brdhmanaih saha bhoktavyam =■ the 
Brahmins must eat together; it is done through ekaSesa. Here 
the root bhuj in bhoktavyam conveys the action of eating in a 
general manner. The verb bhoktavyam must be repeated for 
fully understanding the meaning of the sentence.] 

389. Some point out that major sentences apply 
individually (and collectively). Therefore, they 
include minor sentences, different from one another. 

[The Vrtti points out that the dvandva sentence ordering 
the action of eating quoted above should be split up into 
several sentences like Devadatto bhojyatam, Yajiiadatto bhojyatam, 
Vifnumitro bhojyatam .] 

It is now stated that this does not mean that the major 
sentence is expressive of the meanings of the minor sentences. 

II. 390-392 


390. The minor sentences do not become the 
expressed meanings of the major sentence but when 
the latter is uttered, one understands the other 
meanings of the minor sentences. 

[The Vrtti elucidates the point as follows:—The major 
sentence does not bring about the minor sentences through the 
relation of abhidhdnabhidheyabhava = the relation of word and its 
expressed meaning. The minor sentences which are, after all, 
words cannot be the expressed meanings of the major senten¬ 
ces : Na tasya vdkyasya sabdapadarthakatvam Sakyamabhyupagantum. 
All that is meant is that when the major sentence is uttered, 
the minor sentences, though not uttered and heard, come to 
the mind. Therefore, a minor sentence comes to the mind in 
regard to each of the eaters and thus the meaning of the major 
sentence is applied to each eater:— Tasmdt pratyekam vakyantara- 
prddurbhavah pratyekam vakyasamdptir abhipretd.] 

The upholder of the indivisible sentence criticizes the 
upholder of the individual word. 

391. According to those who believe that the 
whole of the sentence-meaning is contained in that 
of the individual word, of what use then is the exis¬ 
tence of the meaning of the individual word ? 

[If the whole of the sentence meaning is contained in 
the meaning of (any) individual word in the sentence, of what 
use is the meaning of the remaining individual words . . . 
tefdm evambhute pratipadam av as tinted the kimavaSiftendparenapadar- 

thena vikalpitena, says the Vrtti.] 

The upholder ol the indivisible sentence criticizes both 

the possibilities in the opposite view. 

3 92. If the sentence meaning, produced by the 
meanings of the individual words exists in each one 
of them, it is either opposed to it or it co-exists with 


VAKYAPAD1YAM of bhartrhari 

393. If the particularised and the non-parti- 
cularised co-exist, there would be opposition bet¬ 
ween the two. If the meaning (of the individual 
word) is abandoned, then the relation becomes im¬ 

[What is meant here is this : The individual words ulti¬ 
mately convey the sentence-meaning after having first convey¬ 
ed their own meaning. If, once the sentence-meaning is under¬ 
stood, they abandon their own meaning, then the relation 
between the word and the meaning becomes temporary which 
is against the opponent’s own belief. Therefore, according to 
this view, there cannot be real individual words and their 

The Vrtti points out that the meaning of the individual 
word is of a general nature and it can become particularised 
only when its general nature ceases to be : Samanyarupani- 
vrttyaiva hi vi$e$a atmanam labhate. But if the general meaning 
which is first conveyed is abandoned, then the relation bet¬ 
ween the word and the meaning cannot be considered to be 
eternal and that would go against the accepted view of both 
sides:— Atha tupurvam upattarthah {am ?) parityajati nityatvam 
sambandhasya hiyate. ] 

If the sentence attains completion in the collection, is the 
sentence-meaning the meaning of the collection of words or is 
it the meaning of the individual words also? 

394. The one sentence-meaning is the common 
expressed meaning and it is related to each indi¬ 
vidual word. It is related to the collection and to 
each unit in the collection and it is understood when 
all the words are there. 

[The Vrtti here brings in the analogy of the relation 
called samyoga = contact. When two things are united by the 
relation called contact, does this relation exist in each of the 
two objects or in both of them together ? Some think that it 
exists in both ways :— kecin manyante yatha samyogasaiijnd dvayor 

II. 395-396 


dvayoh samuddye cdvatifthate tatha pratyekam dvayoh samuditefu ca 
sa evaiko'rthdtmd vyavaslhita eva. In this connection, the Vrtti 
reminds us of the dvandva compound. When two words together 
form a dvandva compound, they stand for the group and that 
group is expressed by each one of them. Each word in the 
compound expresses the group in which the parts are also 
manifested. Each word conveys at the same time all the things 
meant to be conveyed by the dvandva compound as a whole. 
This principle is called Tugapadadhikaranatd. (See Bhartrhan, 
371 ff) The Vrtti says:— Tatha tar hi dvandvavdcyo bheddnuga- 
V tah samudayah pratibhedam samdpto’bhidhiyate, dvandvasamjnakasya 

samudayasya sa evdithah. . „ 

Samyoga and dvandva were brought in only for the sake of 
analogy The point that has been sought to be made here is 
diat both the sentences as a whole and the individual words 
become meaningful with the meaning of the sentence.] 

An illustration is now given. 

395. Just as all have the feeling of ownership 
towards common property and all enjoy the fruit of 
charity (given out of the common property) and all 
feJl the same amount of joy, in the same way, the 
meaning is related to its words (the individual 
words and their collection.) 

An illustration is now given from the Sdstra. 

396. Even if the phonemes are expressive of 
meaning and the stem having the phonemes also has 
the same meaning, the case-ending expressive of the 
singular number comes after the stem which is a 
collection and not after each phoneme. 

[This verse is based on M.Bhd. I. p. 22o, 1. 10-24. The 
particular sentence in the M. Bhd. referred to here is : Sanghd- 
tasyaikdrthydt subabhdvo varndt = As it is the whole group of 
phonemes which is expressive of meaning, the case-ending is 
not added to (each) phoneme. 



The Vrtti has a gap 5 but the relevant passage is this = 
Samuddyasya tadarthayogad ekalvam, samuddyad ekenotpadyamdnena 
supd vyaktam ityekatvdntardbhdvdt prativarnam vibhaktyutpattau 
nimittam na bliavati. ] 

The same idea is explained by an illustration. 

397. Just as everybody sees common property 
by means of the same lamp, in the same way, by 
means of one case-ending the number is understood. 

[The Vrtti points out that the one case-ending added to 
the stem as a whole expresses the singular number and there 
is no need for another case-ending : Tathaikena supd prak&titam 
ekatvam sarvasambandhigate tasminnarthdtmani pratiyamdnam na 
subantarenabhidhiyate. ] 

398. As far as meaningfulness is concerned, 
theie is no difference between the individual word, 
the phoneme and the sentence. This procedure 
based on practice (tradition) is different and 
appears to contradict the accepted doctrine. 

[It was stated in v. 396 that the case-ending added at the 
end of the stem indicates the meaningfulness, not only of the 
stem but also of the phonemes and of the individual words if, 
an y, included in the stem. This seems to go against the accept¬ 
ed doctrine of the grammarians that only the indivisible sen¬ 
tence is real and expressive of meaning. Really speaking, there 
is no contradiction. It has already been said that the meaning¬ 
fulness of the individual word is accepted for the sake of conve¬ 
nience. But as this question of convenience comes up on so 
many occasions, it appears to be taken seriously but it is not 
so. It is accepted only for the sake of. convenience. Conve¬ 
nience includes also accommodation to other Sastras because 
Vyakarana is sarvapdr$ada = common to all schools of thought, 
as so often repeated by commentators like Helaraja and 
Puny a raj a.] 

II. 399 


The function of the word in conveying its meaning is 
now considered. 

399. Without its application, a word does not 
convey its meaning. That is why it has been declared 
that the relation between a meaning and the word 
is through ukti, that is, application. 

[The following are the points in the Ambakartri on this 
verse. The power of a word to convey its meaning is called 
abhidha. Some consider this abhidhd to be different from 
pratibha, others not. Pratibha when conveyed by the words ukti 
and viniyoga is a vagdharmo, a property of words. When con¬ 
veyed by the words anusandhana and abhisandhdna, it is a mano- 
dharma. When conveyed by the word pranidhana, it is a property 
Qp the vision. Here the word pratibha is used in a sense diffe¬ 
rent from its usual one. It only amounts to the conscious use 
of a polysemic word in one of its meanings. In other words, 
it becomes a synonym of ukti and viniyoga. What is conveyed 
in this verse is that when a polysemic word conveys a parti¬ 
cular meaning in a particular context, it is due to the delibe¬ 
rate application of that word in that meaning by the speaker. 
This is called viniyoga or ukti. 

Two new words are used in this verse: ukti and viniyoga 
and they seem to have the same meaning, namely, the use, by 
the speaker, of a word deliberately in a particular meaning. 
Punyaraja understands them in this way and for him, ukti or 
viniyoga is necessary in the case of words having more than 
one meaning: Iha eka evatabdo bahvarthah, he says at the begin¬ 
ning of his all too brief remark. The Vrtti also seems to speak 
about the deliberate use of a word in a particular meaning 
by the speaker. The relevant words of the Vrtti are: Talrdnena- 
yam vaktavya ityubhayor vdcyavacakayoh parigrahani krtva buddhis- 
thasabdo buddhisthe yatra viniyujyate pravanikriyatc satyapyane- 
kdrthatve tatrasya samarthyam avacchidyate. What the Vrtti says 
amounts to this: when a word has more than one meaning, 
the speaker mentally chooses one of them and decides that it 
should be expressed by the word: anenayain vaktavyah. The word 


VAKYAPAD1YAM of bhartrhari 

in the mind is applied to the meaning which is also in the 
mind: buddhisthasabdo buddhisthe yatra viniyujyate . Then the 
capacity of the word to convey that meaning is brought out 
fully: tatrasya samarthyam avacchidyate .] 

The fact that conscious and deliberate application of a 
word to a meaning is necessary does not make the relation 
between word and meaning man-made. It is eternal. 

400. Just as it is when the eye is directed to¬ 
wards an object that it perceives it, in the same 
way, a word denotes an object when deliberately 
applied to it. 

[The power of the eye to see the object is natural and 
eternal but it does so only when it is consciously directed to 
an object. Similarly, the power of the word to convey its 
meaning is natural but it has to be deliberately applied to it 
especially when it is a polysemic word. 

Punyaraja uses the words abhisandhdna , pratibha and 
pracdra in his commentary on this verse. The word has the 
power called abhidhd but unless the speaker has the desire to 
apply a word to a particular meaning, this abhidhd of the word 
will not function. So some identify the abhisandhdna , the deli¬ 
berate application by the speaker, with the abhidhd of the 
word. This abhisandhdna of the speaker is nothing more than 
pratibha. It is this very pratibha which becomes pranidhdna when 
we direct our eye to a particular object. 

The analogy of the word to the eye is already given in 
verse. While explaining it, the Vrtti concludes : Tathd tabdo'- 
pyanekdrthapratydyanayogyo yamartham pratyabhisamhito bhavati 
tam upasamgrhndti svatmani sannivesayati prakd£ayati = In the same 
way a word, even though capable of conveying many 
meanings, whatever meaning it is directed at by the speaker, 
it embraces it, it becomes one with it and it illuminates it.] 

Something is now said to explain the use of new words 
like viniyoga, ukti, abhisandhdna and abhidhd. 

401. One sees that the relation between the 
object and the instrument is through action. 

II. 402-404 


Similarly, the relation of expression and expressed 
between the word and meaning is through the func¬ 
tion called abhidlia. 

[It is this function called abhidha which is called ukti. 
viniyoga and abhisandhana. The words may be new but they 
denote a well-known function. 

The Vrtti makes the following points:—The accessories 
of an action, like the steps of a ladder, are not directly con¬ 
nected with one another. But they are directly connected with 
the action and through the action indirectly connected with 
one another. In the process of the word conveying the meaning, 
the word is the karana, the meaning conveyed is the karma. 
There can be neither karana nor karma except in regard to an 
action or process: karanakarmanoka kriydm antarena{na) praklptih. 
It is the action or the process which brings about the result 
and not anything else: Phalavati ca kriyaiva , nanyasyarthasya 
phalavattd. The process can be compared to that of unhusking 
grain in which the mortar and pestle are the adhikarana and 
karana respectively and the grain is the karma. Raising and 
lowering the pestle is the action which produces the result, 
namely, unhusking: udyamana-nipdtanarupo , vahantih phalaprasa- 


402. When many things are denoted by a word 
and all of them can be equally well-connected, with 
an action, that thing, which the speaker intends to 
convey, is conveyed by the word. 

[The purpose of the verse is to emphasize the speaker’s 
intention in the process by which words convey their meaning: 
As the Vrtti puts it:— Tatra yat prayoktabhisandhalte tadvacana- 
tvam avyabliicarena sabdasya pratiyate.'] 

403,404. Some scholars who believe in the same¬ 
ness of the word (in all its meanings) declare that at 
the time of the practice (of learning the words of the 
Veda) they are without a meaning, that they stand 


VAKYAPAD1YAM of bhartrhari 

for their form only when they are taught to others 
but that (at the time of the performance of a sacri¬ 
fice) they are expressive of different meanings be¬ 
cause of difference in expressive power based on 
difference of intention. 

\_Niyogabhedat = Abhisandhanabhedat = due to difference 
in intention. Abhidhanakriyabhedat = due to difference in 
expressive power. 

The same Vedic passage is put to different uses and 
therefore, functions differently. As the Vrtti concludes— 
Ekasabdatve hi nimittabhedad ekasyaiva tathd tatha vyavasthd 
many ante.'] 

All this has been said according to the view that a word 
having many meanings is the same word. What particular 
meaning it conveys in a particular context depends upon 
abhisandhana , pratibha , the intention of the speaker and not 
the context itself. 

Something is now being said according to the other view, 
namely, that the word is not the same. There are as many 
words as there are meanings. 

405. Those who believe in difference declare 
that the word is absolutely different in each case, 
like the words aksa etc. though it appears to be the 
same because of the presence of sameness (that is, 
the same sequence of phonemes.). 

[The word aksa can mean a kind of fruit, or playing dice, 
or the axle of a cart and so on; but it is a different word in 
each case, according to this view.] 

406. In such cases,there is no intention beyond 
utterance. As the power of the word is restricted 
to a particular meaning, it is attached to it. 

[In the other view, the intention of the speaker was given 
as the factor which makes a polysemic word convey a particular 

II. 407 


meaning in a particular context. In this view, that factor is 
eliminated. As the form of the word resembles that of others 
having other meanings, only the context can tell us which 
meaning is to be understood. 

The Vrtti points out that undue importance should not 
be given to the speaker’s intention because, sometimes, mean¬ 
ing is understood from words uttered by children who do not 
know the many meanings of a word and who cannot, therefore, 
have the intention of conveying one of them: Tatlid hi balenap- 
ycirtham avidusa prayuktena Sabdenanabhisamhitenapi bhavati frotrna - 
m arth ap ra tip at till. In this view, there is no such thing as one 
word having many meanings. What looks like the same word 
is really a different word. So each word has one meaning only 
which is invariably linked to it: Tasya Sabdasya yo vifayah sa 
tatrdvyabhicdrena vyavasthita eva . As the word may look another 
because of the sameness of the sequence of the phonemes, the 
help of context may be needed to get at the meaning invari¬ 
ably linked with it. 

40 7. From the fact that the meaning is deter¬ 
mined with the help of the context (artha and 
prakarana ), one understands that it is a different 
word. A word which has a certain meaning cannot 
express another. 

[ What is pointed out here is that the very fact that one 
has to have recourse to the context to determine the meaning 
shows that it is a different word : Tat tu tat svabhdvikam asam - 
kirnam arthayattam nityam §abddndm svenarthena nityasambandhan 
nandtvam tatpratip attar am pratisandehanivrttyartham prakamadibhih 
pravibhajyate . The relation between the word and the mean¬ 
ing being eternal and not man-made, a meaning once con¬ 
veyed by its word, cannot be abandoned nor can a meaning 
not conveyed by it be ascribed to it : na tu kadacid updttasya 
svasabdena punastyago ’ sti 3 anupattasya vopadanam . Apauru$eyo 
hyautpattikah fabdarthayoh sa?nba?idhah.\ 

If context can determine the meaning of a word and if 
such a word is different from another having the same sequence 


VAKYAPAD1YAM of bhartrhari 

of phonemes, the established doctrine of the indivisibility 
of the sentence and the sentence-meaning seems to be affected. 
But that doctrine stands. 

408 . These considerations apply to a sentence 
consisting of a single word with the verb asti 
mentally supplied and not to an individual word 
which is part of a sentence (and which has been 
obtained by analysis). 

409 . Just as it is stated by others (the Mlm'am- 
sakas) that the meaningless phonemes manifest the 
expressive word, in the same way, the meaningless 
individual words manifest the sentence having a 
particularised meaning. 

We do seem to understand word-meanings before the 
sentence-meaning is understood. How is that ? 

410 . The cognition of the meanings of the 
individual words which takes place in the middle is 
only a means (to the understanding of the sentence¬ 
meaning) because the sentence-meaning is not 
understood at the beginning. 

[As illustrations of the process of understanding some¬ 
thing in the middle, the Vrtti takes the words bhavati and 
brahmanakambala. In the first case before we understand the 
meaning of the full word we may understand something from 
bhava , but that is not real. In the second word we may under¬ 
stand something from brahmana but in the meaning of the full 
word it has no reality : Tadyatha bhavatiti bhavaSabdamatrasyar - 
tham brahmanakambale ca brahmanasrutimatrasyartham.'] 

The nature of the sentence-meaning according to anvita 
bhidhanavada in now stated. 

411 . Just as (according to abhihitanvaya) the 
meaning of each succeeding word particularised by 

II. 412 


the meaning of the preceding word is the sentence¬ 
meaning, in the same way, the connected meaning, 
present in the very beginning, made clear when 
the meanings of the subsequent words are under¬ 
stood, is the sentence-meaning. 

[It is a connected meaning from the very beginning. 
Each succeeding word-meaning makes this connection clear. 
It is in the nature of an action associated with its accessories. 
The very first word of the sentence conveys this connected 
meaning but not sufficiently clearly. The grammarians’ view 
is that it is in the nature of intuition, without inner sequence 
and indivisible. 

The Vrtti makes the same idea clear as follows— Tatha 
samsargasya prakrantatvat prathama eva Sabdah faftyartham upadaya 
pravrttah, sa tu nityo viSifto nitya eva padantarasannidhanat prati- 
pattrfii vyaktim labhate .] 

412. According to some, once the need for the 
accessory is accepted because of the action to be 
accomplished, the accessories are again actually 
mentioned in order to specify the substrata of the 
power of the accessories. 

[Those who hold the anvitabhidlidnavada are referred to 
here. If, from the very beginning a connected meaning invol¬ 
ving an action and its accessories are understood, why the 
accessories are again mentioned in the sentence is explained in 
this verse. It is for specifying the substrata of the powers of 
the accessories. 

The Vrtti points out that action is a process, something 
to be accomplished ( sadhya) and the accessories which help in 
the accomplishment are already there ( siddha) and they exist 
for the sake of the action ( pardrtham apadanat ) and they are 
Se$a (secondary) in regard to the action ($e§abhdvemiigi- 
krte$u sadhanefti ) and they assume the form of the action so to 
speak ( kriydrupa-manupravistefviva ). This connection between 


VAKYAPAD1YAM of bhartrhari 

action and the accessories is understood from the very begin¬ 
ning and if the accessories are actually mentioned, it is in order 
to make clear what the substratum of the power of the acces¬ 
sories is : adharapratipattyartha dravyasrutib. ] 

413. As the substratum is not specified it is 
not understood. In the nature of things, there is 
the possibility (of a specific substratum) and the 
actual mention is for excluding others. 

[The Vrtti says :—An action may bring to the mind all 
the possible powers which can help in its accomplishment 
but not any particular substrata of these powers : kriyd sam- 
bhavinah saktivisesan sarvan akfipati, naivam asya dravyavisefdpckfd 
pratibandho rupalabhe. But it brings the substratum of the 
power of the accessory in a general manner to the mind and if 
any particular substratum is mentioned in the sentence it is 
in order to exclude others : Dravyasdmdnyam tu sadhanadhara- 
tvena samarthyam grhitam. Tatra visepasrutir ddhdrdntaranivrttyarthd 
pratiyate, says the Vrtti. ] 

414. An action with specific accessories as its 
substrata and (therefore) different from other 
actions, is understood from the very beginning and 
its specifications ( bhedah , mentioned in the sentence) 
are for the sake of the understandingof the listener. 

[This is also anvitabhidhanavada. ] 

415. According to others, the sentence and the 
sentence-meaning, indivisible and devoid of the 
sequence of the words shine from words having a 

[By anyefam the grammarians are meant here. They 
include, of course, Bhartrhari himself. Another example of 
Bhartrhari giving his own views as though they were those 
of others. ] 

II. 416-417 


According to Punyaraja, the purpose of the following 
-verse is to state that the meaning of the real word, that is, the 
sentence, is the real conditioned by the unreal. 

416. That which has a form, its essence is 
indefinable. That which has no form, it is the 
essence of that which is definable. 

[ Couched in very general terms the meaning of this 
-verse is not too clear. Punyaraja understands the first half 
as relating to the meaning of the individual word and the 
latter half as relating to the meaning of the sentence. The 
meaning of the individual word is said to have a form svarupam 
.vidyateyasya) because the object meant by it can be perceived. 
But it does not thereby become fit for verbal usage. That is 
why it is said to be indefinable (tasyatma na nirupyate). The 
meaning of the individual word, isolatedfrom the meanings of 
the other words, especially the verb, is unfit for verbal usage. 
It is at best a means to the understanding of the sentence¬ 
meaning, the essence of which is interconnection of the mean¬ 
ing of the individual words and which is, therefore, fit for 

verbal usage and thus real. 

It is a pity that the Vrtti on this verse is obscure. It has 
gaps in several places. As usual, Punyaraja’s commentary seems 
to be based on it. Sentences like taccdsam vedyani vyavahardtitam 
in the Vrtti have obviously influenced the wording of Punyaraja’s 
commentary. But one cannot ignore the strong impression 
which one gets while reading the Vrtti that its analysis is deep¬ 
er and that it contains some points not found in Punyaraja at 
all. Unfortunately due to the unsatisfactory nature of the 
text, it is not possible to note down those points.] 

417. Others think that the meaning (of a sen¬ 
tence) cannot be determined through the word. A 
remembrance resembling the experience of the 
object takes place through the words. 

[Some think that words are not the means of understand- 


VAKYAPAD1YAM of bhartrhari 

ing the meaning of a sentence which is in the nature of an 
interconnection ( samsarga ). According to them, this 
interconnection is understood by means of an integral cogni¬ 
tion by the mind, ( nirvikalpakaikasamadhigamyam —Punya- 
raja) and not through the meaning of individual words. 
Individual words do no more than cause a remembrance 
similar to the experience of objects. Thus they are too far 
away from the sentence-meaning. This is the gist of Punya- 
raja’s commentary. 

The Vrtti comments on this and the next verse together 
more on the next verse than on this one. The only point 
which it mentions relating to this verse is that the idea which 
we get of an object from its word is far removed from the real 
nature of that object. The next verse explains this very point 
by means of an example : Sarvatrasabdam arthanam svabhava - 
vadharanam , sabdavrttam tvanupatad duribhavati tasmad ityeke$am 
darSanam = The understanding of the nature of an object takes 
place everywhere through other means than the word, what 
the word brings is far removed from them, such is the view of 
some. ] 

418. One who gets burnt experiences the burn 
through contact with fire in one way and one 
experiences it in another way through the word 
burn ( claha ). 

[The word artha in Sanskrit denotes both the external 
object and the ‘meaning’ of the word for that external object. 
The big difference between the two is pointed out here. As 
the Vrtti puts it : Anyatha hyagnihimasastradisannipatad dahadayo 
, vast hah pratyavabhasante. Anyatha dahadibhih Sabdaih pratydyyante. 
The rest of the Vrtti is unfortunately not clear.] 

419. Just as the senses which have each their 
own essential nature and their own field of opera¬ 
tion, cannot perform their function without the 

420. In the same way, the individual words 

II. 421-423 


though expressive of their own meaning, have no 
meaning at all if they are isolatedfrom thesentence. 

[Punyaraja says that the understanding of the meaning of 
the individual word is an error and as such it is either 
viparitakhydti or asatkliyati , two well-known interpretations of 
error in Indian philosophical circles. In any case, if they 
have any meaning at all, it is only as long as they are within 
the sentence and not in isolation. This is what the Vrtti also 
emphasises in its concluding sentence :— Tathaiva prthagarthand- 
mafiipaddnam vdkyopanibandhanatdm antarenartha pratydyanasaktir na 
vidyate .] 

421. The meaning in the nature of an inter¬ 
connection is grasped when the things are connect¬ 
ed together. (When the word meanings are in 
isolation) the essence (of the sentence-meaning) 
does not become clear because it is seen to be diffe¬ 
rent from the word-meaning. 

422. Even knowledge (or consciousness) does 
not appear in its true (pure) state. It is formless 
but appears as coloured by the object. 

[Knowledge or Consciousness is brought in here for ana¬ 
logy- J ust as P ure consciousness is formless but always appears 
as coloured by the form of some object or other, in the same 
way, the sentence and the sentence-meaning are indivisible and 
undifferentiated but appear as the inter-connection between 
words and meanings. The isolated word and word-meaning 
are unfit for communication and that is why they are said to 
be unreal ( asatya ).] 

423. Inasmuch as even the meaning of a single 
word is expressed as connected with existence or 
non-existence, it is the sentence which is used (for 


VAKYAPAD1YAM of bhartrhari 

[Only the sentence and its meaning are real because 
only they are fit for communication. That is why, when a 
communication is made by means of a single word, it is com¬ 
pleted by adding mentally the word or atleast the idea of 
existence or non-existence. 

The text of the Vrtti is doubtful at the beginning but be¬ 
comes clearer later. After having said that like existence or 
positive activity, non-existence or absention from activity is 
conveyed by the sentence. That is why it is the sentence with 
the verb actually present in it or inferred which is used in all 
communication :— Pravrttivacca nivrttisamsargo’pi vakyadharma 
eva. Tasmdcchrvyamdna-kriydpadom anumiyamanakriycpadcm id 
vakyam eva sarvavyavaharefu avatar ati.~\ 

424. No word-meaning, whether real or unreal, 
is understood in communication except as connect¬ 
ed with some action. Therefore, it does not really 

[Explaining this verse, the Vrtti says that one cannot 
predicate the truth or otherwise of the meaning of a single 
word: . . . kevalapadaprayoge satyatvani viparyayo va na prakhyayate. 
It is only when it is completed by a verb that the listener under¬ 
stands it as true or untrue : kriyapadopasamhare tusatyasatyabha- 
vena pratipattrfu vyavaharo vatif thate . Without connection with 
some verb, the completion of the meaning can be done in one 
of many possible ways, in a way that would come under any 
one of the six transformations of Being = sadbhavavikdrdh. That 
is much too vague : so rthah pariplavamano’pi hy asau fad bhava- 
vikdraparydyendnu dhavatiti. That is why, the isolated word-mean¬ 
ing is beyond the scope of communication and is said to be 
unreal : Tasmat kevalapadartho vyavaharatitatvan nastiti vyapa- 

425. Even a sentence having the form of the 
single word sat ( = it exists or existent) cannot be so 
understood without connection with some word 
expressive of action such as ‘it was’ ( tad abhut) , or 

II. 42 6-427 


it is ( tad asli ), or it was not (tan na bhul ) or it is not 
( tannasti ). 

[The point here is that action and accessory require each 
other. So where only one is mentioned, the meaning is in¬ 
complete and the other is mentally supplied to complete the 
meaning. Once that takes place, there is a sentence which is 
fit to be used for communication. 

Without some process or action being mentioned or 
mentally supplied, what an isolated noun brings to the mind 
is too vague and not firm says the Vrtli—Alrdpi sadhyabliidhanam 
antarena pariplavamdna ivdrtho na vyavatifthate.] 

42 6. The requirement (that is, the incomplete¬ 
ness) which is felt in the meaning expressed by the 
verb which depends upon an accessory is not re¬ 
moved unless the thing which can be the accessoiy 
is also mentioned. 

42 7. It is the action which is first analysed 
from the meaning (ofthe sentence) because it is the 
main thing. The accessories are employed for the 
sake of the action to be accomplished. As for the 
action, it is the result aimed at that sets it in 

[ If the action requires the accessories and if the accesso¬ 
ries require the action, there is equality in dependence. Why 
should the dependence of the nouns expressive of the accesso¬ 
ries be emphasised ? ydvatdyathaiva ndmnam kriydpadam akanksam 
vicchinatti tathaivdkhydtdndmapi vim sattvdbhidhdnapadendkdrm na 
vicchidyate, asks the Vrtti and answers as follows —Sadhyastv 
arthdtmd svaphalaprayuktah prddhdnydt sarvasya vdkyopasangrahar- 
thasya pun am pravibhajyate= The process or action, set in motion 
by the result to be attained is the most important element and 
it is, therefore, analysed out of the sentence-meaning first. 
Once that is obtained by analysis, the accessories on which 
depends its coming into being, are automatically understood. 



Tcna tu pravibhakte^jia ) sadhanapralabhyatvad, atmalabhasya 
samarthyak$iptani sadhanani pratiyante. This is not the case with 
the accessories which are all already there (siddha) and are 
set in motion by the process (sadhya prqyuktefu—Na tu sadhanefu 
siddhatvat sddhyaprayukte$v etat sambhavati .] 

It is now stated that all this conception of process and 
accessory and their mutual relation is artificial and relative 
and not real. 

428. It is the speaker who thinks of something 
as a process and something else as its accessory and 
it is he who thinks of the relation between the two. 

[The point here is that there is nothing fixed about 
these three things: that which is to be accomplished, that 
which helps in its accomplishment and the relation between 
the two. It is a matter of the speaker’s choice. As the Vrtti 
puts it—Togyasaktinibandhanaya vivakfqyd Sabdapramdiwkah pra- 
yoktd iabdasaktim anugacchanu api vivaksantam tainarlhani tat/id 
tatha samihate. Pfa hi vustutdli sadhyatvam sxddhatvarri vd sabda- 
pravrttinimittam. Kihcavyavasthitam aniyatam vidyate.] 

If these were fixed, nobody would be able to change 
them. But we see that the same thing is presented differently 
by different speakers. 

429. The action of cooking is expressed as 
the object in the sentence : I do the act of cook¬ 
ing (pacikriydm karomi). But the actual doing (of 
the act of cooking) is understood as a process 

[The point here is that the action of cooking, taken as an 
example, is presented now as a process and now as the acce¬ 
ssory ( sadhana) called karma= object of action, according to 
the wish of the speaker. There is no pre-determined fixity in 
this matter. The Vrtti also emphasises this very point : Pacikri¬ 
ydm karotityubhayor avisifte sddhydtmakatve paktih sadhyatvena 
vivakfita siddharupena sadlyariipetia vabhidhiyate. It is not clear 

II. 430-433 


Avhat is meant by ubhayoh . The reference may be to the two 
actions mentioned in paeikriyam karoti : (1) The action of cook¬ 
ing expressed by the noun paeikriyam and (2) the action of do¬ 
ing expressed by the verb karoti . The former is presented by 
the noun as a thing ( siddha ) and the latter as a process 
(sadhya) by the verb. Apart from presentation by words, an 
action is really a process and it is so meant in karoti , : As the 
Vrtti puts it : karanam tu sadhyatvenaiva vivakfitatvdd (vi ?) pra- 
krtdvasthdm na jahdti .] 

430. As an object lias all powers, it is present¬ 
ed as rendering whatever service the speakers 
think of. 

[An object is as the speaker sees it and wishes to present 
it in words and so it is presented differently by different speakers. 
The simple act of cooking can be presented differently as in 
the following expressions : pakam karoti, pacati, pakasya nirvrttim 
karoti. Irrespective of what it is in reality, an object follows 
the intention of the speaker and appears to be as he presents 
it. As the Vrtti says : Vastu-sambandham anapeksamdnah prayo - 
ktur vivakfam anupatams tathaiva vyavatif (hate .] 

431. Sometimes objects which are far from one 
another are presented as being connected and those 
which are near one another are presented as be¬ 
ing apart. 

432. Separation of what is united and union 
of what is separated; what are many are presented 
as one and what is one as many. 

How the same thing is presented in different ways is now 

43 3. Because every object is everything (as it 
has all powers) or because it has no essence at all. 
This can be explained. In all this, the word, the 
power of which is extremely restricted, is the basis. 




[In the previous verses, the author has been speaking 
about the great variety and the consequent lack of fixity in 
the way in which objects are presented by words. It has been 
pointed out that much depends upon the intention of the 
speaker and his ability to make use of the capacity of words 
to present the same thing in different ways. In this verse, two 
alternative ways of looking at the matter are mentioned. In 
the first way, the object is endowed with all powers, it is sarvat- 
maka but the speaker, urged by his own purposes and inten¬ 
tions, cognises and determines some aspect of it and presents 
it through words which have the capacity to express that aspect. 
In the alternative way, the object has no essence of its own 
(nairdtmyat ). It is as the speaker cognises it- and expresses it 
through words whose expressive power is fixed. This is the 
gist of what Punyaraja says. 

He is only putting in his own words what the Vrtti had 
already said before in what is to us to-day a more obscure lang¬ 
uage. It begins by saying that the sentence-meaning is very 
comprehensive and includes everything that the speaker might 
want to convey and has all powers. Or it might be looked 
upon as having no essence at all. The Vrtti says much about 
the speaker’s intelligence and intention in all this, but what it 
says is not too clear]. 

434. The word is only an adventitious mark of 
an object ( upalaksana)] it does not express the 
service which it renders, it is not capable of touch¬ 
ing (that is, of expressing) the powers of the 

[The point here is that a word really does not touch the 
essence of an object. It stands far from it and does no more 
than somehow bring it to the mind, so that it can be talked 
about. There is no real relation between the two. 

The Vrtti , following, of course, the verse, also emphasises 
that the word can at the most denote the object, that is, bring 
it to the mind but it cannot express the powers of the object. 
It is the powers which distinguish an object from others and 

II. 435-437 


enable it to render service, but the word is not capable of 
denoting all that :— Sa hi vastumatrasamsparSitvad bhedakanyupa- 
kdrini saktirupani na samsprSati.~\ 

Even the relation of expression and expressed ( vdcyavaca- 
kabhava) between word and meaning cannot be maintained. 

435. The relation called contact is expressed 
as that which is related ( sambandhin ) by its own 
word (namely, samyoga ); similarly, inherence is 
also expressed as that which has inherence. 

[ In the expression; bhuiale ghatah= the jar is on the floor;, 
the relation of contact between the floor and the jar is under¬ 
stood as a relation but in the expression dravyayoh samyogah= 
‘the contact between the two substances’ where the word sani- 
yoga (contact) is itself used, the relation is not presented as a 
relation but as a sambandhi= that which is related. In other 
words, its own word does not present it in its true nature. It 
presents it as the related whereas it is really a relation. The 
same is true of the word samavaya which means the relation 
called inherence. ] 

436. Objects are presented in some capacity 
or other and not in their real nature. One and 
the same object is understood as the basis of some 
service or other. 

[Words never present an object in its true nature. They 
pick on some capacity or function of it and present it in that 
form. So what the word presents has no fixity. It is not real. 
It is always relative. The Vrtti ends its comments by bringing 
in the illustration of the same woman being daughter, sister, 
wife or mother in relation to different persons. Tadyathd ekd 
stri duhitdbhagini blmryd mdtetyapekfdviSefaih pravibhajyate.] 

It is now stated that the sentence-meaning, being always 
new, is real. 

43 7. The essence of the sentence-meaning 
which is in the nature of inter-connection does not 


VAKYAPADlYAM of bhartrhari 

rest anywhere. In communication ( vyavahare ) it is 
the soul of the word-meanings. 

[The idea here is that the sentence-meaning, like know¬ 
ledge or consciousness is difficult to be grasped except as 
coloured by the word meanings. As it connects together the 
word-meanings and makes them fit for verbal usage. It is 
their soul. It exists in the meaning of each word or in all of 
them together. 

The Vrtti points out that the meaning of the individual 
word is of a very general nature and it is only when it is 
connected with the meanings of the other words that it becomes 
particularised. The particularisation of the meaning is the 
chief feature of the sentence-meaning and even if one accepts 
division within the sentence, one has to admit that this parti¬ 
cularised meaning exists in the meaning of each division i Tatra 
bhedadarsanam asritya kaiscid ucyate pr atipadarthain tasyatmd vyavas - 
thita iti . ] 

438. In reality, it does not rest in the word¬ 
meaning or in all of them together or anywhere 
else. However its essence is differentiated on the 
basis of the artificial differentiation in word mean¬ 

[The form of the sentence-meaning as distinct from that 
of the word-meanings cannot be indicated. Its form is deter¬ 
mined in terms of the word-meanings. In its own pure form, 
it is beyond worldly usage. 

The Vrtti points out that when the individual words are 
used, particularisation of meaning takes place and from this 
particularisation, one infers that a power, different from the 
power of individual words to convey their general meaning, 
exists which is responsible for the particularisation : Tasya tu 
kevalapadaprayoge yo vUe?o nirdharitah tena vi§e$enanumanena sam- 
bandhat sa prthak faktih pravibhajyate. ] 

439. The division of it which is made in order 

II. 440-441 


to explain it is the means of knowing it. In that 
division, the parts are incomplete and require one 
another. Through this means, something different 
from it is defined. 

440. Of the sentence which has many powers, 
a division is understood (on the basis of the diffe¬ 
rentiation of word-meanings). The unity of the 
sentence-meaning is understood even from slight 

[For the understanding of unity from slight indications, 
Punyaraja gives the following illustration. There is the siitra 
lg yanah samprasaranam (P. 1.1.45) =The ik which comes in 
place of yan is called samprasaraija. The two words ig yanah 
convey a connected meaning, namely, the relation of original 
and substitute (sthanyddetabhava) . It is a sentence-meaning 
because it comes from two connected words. But this sentence 
meaning can be conveyed by the single word in the singular 
number, namely, samprasaranam . This is the slight indication 
( mdtraya ). 

What the Vrtti says may be briefly, indicated as follows:— 

Those who uphold the unity and the eternality of the 
word believe that the sentence conveys action, particularised 
by time, special accessories, substance, person, aspect etc. 
That One indivisible word is analysed for practical purposes 
into parts, but it stands for one indivisible meaning, involving 
qualifying and qualified elements. By analysing the unified 
powers of the sentence, practical divisions are obtained. But 
without this practical division, words like samprasarana are con¬ 
veying a sentence-meaning. The relevant words of the Vrtti, 
though doubtful in places, are as follows— Tena vakyarthastvekah 
samprasdranddiprakdraya matraparimanayd vibhagoddesam antarena 
pratyayyate . Vakydrthasya hi samprasdranasamjhdvibhdgoddekna 
vind sambandhini vijiiayate . ] 

441. From the sentence-meaning which is of 
the nature of cognition, a word meaning having or 



not having something external corresponding to it 
but appearing to have it is analysed. The analysis 
consists in abstracting the powers capable of con¬ 
veying the word meanings. 

[This verse says that the sentence-meaning is something 
mental, not having anything corresponding to it in the external 
world. As it has nothing corresponding to it in the external 
world, it has not been seen and that is why it is looked upon as 
something new. It is in the nature of a connection and connec¬ 
tion is something mental. That is why it cannot be a real 
part of the sentence-meaning. Anything which appears to be 
a partis unreal. As the sentence is mental, its parts, if any, 
can also be only mental, but it appears to be external or it is 
identified with the real external object.] 

441 (a). The powers relating to the mental 
objects ( pratyayarthatmanah ) are not clearly deter¬ 
mined. Nor are their forms obtained elsewhere 
(than in the sentence-meaning). 

This verse is not found in Punyaraja’s text but it is found, 
in some manuscripts of the verses only and it is there in the 
only manuscript of the Vrtti which shows that it is not a recent 
interpolation. It is older than Punyaraja, for some reason, 
the manuscripts of Punyaraja’s commentary, utilised for this 
edition, do not have it. I have included it as it is authenti¬ 
cated by the Vrtti . As it is connected in meaning with 441 and 
in order not to disturb the numbering of the following 
which agrees with R and RP, I have numbered it 444(a) verse. 

The Vrtti on this is not too clear. In one place, there 
is a gap also. It seems to point out the difference between the 
external object and the idea of it which figures in the meaning 
of the individual word. The latter is only an imitation of 
the former and yet that is all which we have in order to under¬ 
stand the sentence-meaning. Through it a division is made in 
the sentence-meaning and this division is the means of under¬ 
standing the sentence-meaning : Tatra bahydrthavi$ayanam 

II. 442 


laktlndm abhavad anukaramatram upadayabhedah prakramyate. Ayam 
xvahi bhedah pratipatterupdyah. 

What is said in 441 and 441 (a) may be briefly stated as 
follows, though the idea is not too clear—“The meaning of the 
individual word, understood as a constituent of the sentence¬ 
meaning, is something mental only, but it appears to be exter¬ 
nal. There is, of course, the external object, but it figures in 
the mind as the meaning of the individual word. Then it is 
only mental and it appears to be identical with the sentence¬ 
meaning which is wholly mental. After the sentence-meaning 
is understood, it is analysed out of it. The powers of the sen¬ 
tence meaning, favourable to the understanding of the word- 
meanings, are analysed out of the former. These powers are 
not clearly determined. According to the view that the sen¬ 
tence-meaning is mental and indivisible, it can have no powers 
leading to division. How can they be then analysed out of it ? 
If they exist anywhere, it can only be in the external object 
and they are only inferred. Without looking upon the external 
object and the one figuring in the mind as endowed with 
powers, worldly transactions cannot take place. The latter is 
identified with the former and this is what is meant by bdhyi - 
krtya in 441. Because of the identification of powers, this ana¬ 
lysis of the mental sentence-meaning is possible. If the mental 
object, identified with the external object, is unreal, then the 
external object is real. Otherwise, the mental object is real 
and not the external object.] 

Another question is raised. When there are several phra¬ 
ses, each having a verb and each incomplete and requiring the 
others, do they make up one sentence or should they be looked 
upon as many sentences ? 

442. Even when there are many verbs 9 they 
constitute one sentence, if they require one an¬ 
other. In this way, the prohibition of loss of accent 
after verbs would be meaningful. 

[There is a reference here to the difference of opinion 
between Panini and Katyayana. According to the former, 



there can be more than one verb in a sentence, provided one 
of them is the main one and the others, subordinate to it. 
According to the latter, more than one verb means more than 
one sentence. See M. Bha. III. p. 373, 1. 10 and p.374, 1. 

443. According to him who holds that the 
fixed definition of the sentenced that it should have 
only one verb, there would be a different sentence 
if there were more than one verb and, therefore, 
the mention of atihah would be useless. 

[According to the author of the Vdrttikas , there shouldT>e 
only one verb in a sentence. More than one verb would result 
in more than one sentence. So the other words in the sentence 
should be non-verbs. In P. 8.1.28, a verb coming after a non¬ 
verb is said to lose its udatta accent. Loss of udatta accent can 
take place only if the non-verb and the verb are both in the 
same sentence. If the sentence can have only one verb, there 
is no need to say coming after a non-verb ( atihah ) in P. 8. 1. 
28, because the other words in the sentence are necessarily non¬ 
verbs. So the verb would come necessarily after a non-verb 
or after nothing at all if it happens to be the first word in the 
sentence. If there is another verb in the sentence, there cannot 
be loss of accent at all. Thus the mention of atihah in P. 8.1. 
28 serves no purpose according to this view. ] 

According to some, even where there are many 
verbs, there may or may not be difference in sentence depen¬ 
ding upon circumstances. 

444. In sentences like mrgah pa'syata^ yati ^Look ! 
the deer is going 5 . Where there is more than one 
verb one being connected with another which is 
already connected, there cannot be unity and 

[The sentence mrgahpasyata yati is understood differently 
by different people. First mrgah is connected with yati and that 

II. 445 


makes one sentence, meaning c the deer goes’. Then pafyata is 
connected with mrga, changed to mrgam with the meaning of 
ydti transformed into an adjective. That makes the second 
sentence yantam mrgam paSyata look at the deer that is going. 
In each sentence there is a verb with its own accessory, agent 
in the first sentence and object in the other. This is one 
way of understanding this sentence. 

The other way is : mrgo ydti makes one sentence and 
that sentence becomes the object of the verb pasyata and as 
ydti and pasyata require one another, the whole thing remains 
one sentence. Thus, in the first view there is bheda and in 
the other view there is abheda. In spite of this, one can hold 
the view that there is no difference of opinion between the 
sutrakara and the Vdrttikakdra. When the author of the varttikas 
says that there should be one verb in a sentence, all that he 
means is that there should be only one main verb. The pre¬ 
sence of other subordinate verbs will not lead to Vakyabheda = 
plurality of sentences. 

All that the Vrtti does here is to say that there would be 
loss of accent ( nighdta) if the whole thing is looked upon as one 
sentence and that there would be no nighdta if one sees two 
sentences in it.] 

The point now to be discussed is : when is a meaning to 
be considered complete and when incomplete. 

445. Where the mode of performance of the 
action (denoted by the sentence is desired to be 
known because of its very nature (samarthjlat ), that 
sentence containing no verbal incompleteness is said 
to be complete in meaning. 

[The point here is that the incompleteness must come out 
of the words themselves before the sentence can be considered 
incomplete. When one says : ‘Devadatta cooks’, the sentence 
is complete, though how and what he cooks is not mentioned. 
From the word ‘cooks’ an average listener will understand a 
complete meaning. But when one says ‘Devadatta cooks his’,, 
one feels at once that the words themselves are incomplete. 



The Vrtti gives the sentence vrihayo'vahanyantam as an exam- 
pie of a complete sentence though no details are mentioned in 
it. It is complete because there is no verbal incompleteness . 
sabdalaksampekfayogdd iti, as Punyaraja says. ] 

446. Where there is a bare statement of fact 
and its implications are understood from its express¬ 
ive power ( sruteh ) without the use of words to 
express them, then the sentence is complete. 

[ In other words, the incompleteness must come from the 
words themselves. Otherwise, the sentence must be considered 
complete, even if the listener understands more than what the 
words actually say. The words used do not require other words 
for yielding a complete meaning :— parisamaptameva sabdatii- 
bandhandya dkankfdyds tatrabhavdd vdkyam vijndyate. ] 

Two sentences may say the same thing, and yet there 
may be difference of emphasis in the two statements. 

447. ‘Study while moving around’ ‘move 
around while studying’. In these two sentences 
what is enjoined is the same btit there is a diffe¬ 
rence due to difference in verbal presentation. 

[The purpose of both sentences is to tell somebody to 
study or pray while moving about. The first sentence presents 
the study through a verb and the moving about through the 
present participle and the second sentence does it vice-versa. 
What is presented through a verb is the more important ele¬ 
ment verbally. But in reality the purpose of the speaker is 
the same in both sentences, namely, to tell somebody to study 
(adhi) or to pray (jap).] 

448. Varieties of action, having their fruit and 
involving many minor actions and infinite in 
number because of different modes of performance 
and different aims are expressed by the same verb. 

II. 449-452 


[ It is stated in this verse that though the verb denoting 
he action may be the same, the action may differ because of 
•diffeience m the mode of performance and in the results ] 

449. The verb always denotes an action with¬ 
out any inner variation. From the mere word, all 
the variations cannot be imagined. 

450. In the sentences : ‘the kings will perform 
the asvamedha sacrifice ‘the Brahmanas are holding 
a saciificial session’, the variations are not under¬ 
stood because of difference in the verb. 

[In 4o0, two distinct verbs are used :yaksyante and sattram 
asate but it is not because of difference in the verbs that 
difference in the ritual is understood but because of inherent 
difference in the actions themselves ( arthasamarthydt .) 

Commenting on 449, the Vrtti points out that variations 
in the action conveyed by the verb are brought about by 
differences in the sadhana, the accessories to action ( sadhanablie - 
dena tadbhedasya caritdrthatvdt). The verb itself does not convey 
the variations. Commenting on 450, the Vrtti, the text of 
•which is doubtful in several places, points out a difference in 
the action denoted by yaj in yaksyante and that denoted by as 
in sattram as ate. ] 

451. The action in regard to the seventeen 
prajapatyci animals, though mentioned only once 
and not repeated, becomes diversified because of 
the nature of things. 

[ By samarthyat, what is meant is that, as there are seven¬ 
teen different animals, the same action necessarily becomes 

diversified, though the word denoting the action is not 
repeated. ] 

452. The action of eating relates to Devadatta 
etc. separately. Or the sentence itself is applied, 
to each agent ( pratisvatantram ) separately. 


VAKYAPAD1YAM of bhartrhari 

[ In the sentence Devadatta-T,ajiiadatta-Vipiumitrd bhojyantdm 
the meaning of the sentence is the action of feeding which is 
connected with Devadatta etc. separately. That is what the 
first half of the verse says. The second says that alternatively, 
the sentence itself is split up into three different sentences, 
each having one of the three agents as the subject and the 
verb bhojyatdm as the predicate. According to the second alter¬ 
native, one verb becomes many verbs and therefore, one sentence 
becomes many sentences. As the Vrtti puts it— Pratikartr ca 
vakye bhidyamane bahavah kriyasabda upaplavante. Tasmadekopanib- 
andhanani bahuni vdkya.nyupa.jayante'. If a separate sentence is 

made for each agent, many verbs would arise and thus many 
sentences based on one sentence would come into existence.] 

453. At the time of utterance of the sentences, 
a separate form is not heard but at the time of the 
understanding of their meaning, the form of 
different (sentences) is understood. 

[ In the sentence under discussion, only one verb denot¬ 
ing one action is heard, but at the time of understanding its 
meaning, the verb or its meaning is construed with each agent 

The Vrtti points out that according to some, it is in the 
nature of words that they are uttered in one way and under¬ 
stood in another way: Sabdapravrttidharma evayam anyathoccaranam 
anyathapratipattih. At the time of understanding the meaning 
of a sentence, it becomes diversified when many sentences arise 
out of it:— Pratipattikale tu vdkydntare$upajayamane$u pravibhaktam 
rupantaram pratiyate .] 

This happens in the kas Ira also. 

454 . An all-embracing sentence (grahanakam 
vakyam) like kartari ( krt ) (P. 3.4.67.) is enunciated 
in a general manner and then it is applied sepa¬ 
rately to cases like pasu. 

[P.3.4.67, says that a krt suffix is added to a root in the 
sense of the agent of the action. This general statement is applied 

II. 455-457 


to particular cases as in P.3.2.25 according to which the krt suffix 
in is added to the root hr when the preceding proximate word 
( upapada ) is drti or ndtha and the agent is an animal. Thus we 
get the form: drtiharih (pasuh ). If the agent is not an animal, this 
suffix cannot be added to the root. The form would be drtih- 
arah with the suffix an (P.3.2.1.] 

455. If, once the general rule is uttered (with 
some special rule) and its incompleteness is remov¬ 
ed, it would not be connected with the other 
(special rules). 

[And so the required forms cannot be derived. The idea 
here is that a general rule should have some scope as a general 
rule, apart from its scope when associated with some special 
rule. Otherwise, some forms cannot be derived. The general 
rule, must not be tied up with any particular special rule, as 
that would take away its scope as a general rule as well as the 
possibility of its being associated with other special rules. As the 
Vrtti puts it:— Tadi tu kartari krdityekasminneva s utre paivadibhir 
akdnksa vicchidyeta samanyena sadhutvam na syat. Ekena vopadhind 
krte’vaccliede tatraivavaruddhatvad upddhyantarasambandho na prak- 

456. It has the same form, it has many appli¬ 
cations and, therefore, it is the basis ( upanibandhana ) 
(of the subordinate rules); it is the source of the 
rules of elaboration ( vibhagavakyanam) and it appears 
to be one with them. 

[A new point of view is discussed: how can action which 
is something to be accomplished and has parts arranged in a 
temporal sequence be looked upon as a universal and as a parti¬ 
cular ? The next verse tries to answer the question.] 

45 7. Sometimes action renders service through 
its individual aspect and sometimes it is its general 
aspect which fulfils some purpose. 



[How can action which is a process ( sadhya ) and has parts 
arranged in a temporal sequence be looked upon as the universal 
or the particular ? It is like this: where the accessories are men¬ 
tioned, in prohibitions, in optional instruction, in accumulation 
in expressions of excellence and praise, it is its individual aspect 
which counts. Where the accessories or time and place are not 
mentioned, it is the general aspect that counts. 

458. Where variations are due to time or in 
expressions like ustrdsika, it is the general aspect 
that works and they do not affect the form of the 

[This stanza states where the verb denotes the general 
aspect of action.] 

It was stated before how many actions conveyed by the 
same word are connected with different accessories conveyed 
by different words. Now how many actions conveyed by 
different words are connected with different accessories con¬ 
veyed by the same word is stated. 

45 9. Where actions differ from one another in 
their universals etc, (and are conveyed by different 
words) and the accessories are also of the same 
number but are conveyed by the same word uttered 
once, the former are separately connected with the 

[In ak$a bhajyantdm , bhujyantam, divyantam , the three 
verbs denote three different actions unconnected with one 
another and are conveyed by three different words. The word 
ak$ah conveys three different accessories though it is uttered 
only once by the speaker. But the hearer connects each 
accessory with one of the verbs and thus gets the complete 
sentence. ] 

460. Different actions like breaking, eating and 
playing are connected separately with the accesso- 

II. 461-465 


ries, namely, the aksaseve n though the latter are 
conveyed in one utterance and simultaneously. 

461. In the case of the three kinds of aksas- 
conveyed by the same word, the use of the same 
word is a means (of conveying them at the same 
time). When they are separately conveyed, there 
is sequence; the single utterance is for conveying 
them at the same time and connecting them 

462- There are two ways of presenting things 
through words : sequence or simultaneity. The 
world does not go beyond them. 

463. Where words are used in a sequence, 
their form varies but not when there is simultaneity. 
Even when there is simultaneity (in the presentation 
of the accessories) the action follows sequence. 

[To say that the action follows sequence means that each 
action is connected separately with the different accessories.] 

464. Difference and unity are two powers 
which appear 1o be different from the word. Even 
where many (accessories) are conveyed at the same 
time by the same word, the latter becomes diversi¬ 
fied (when its many meanings are construed) in 
usage (with different actions). 

[Though the word aksah is one word which conveys many 
accessories at the same time, it becomes many when it is con¬ 
nected with the different verbs, each conveying a different 

465. Where a complex whole with its parts in 
a patent state is meant to be conveyed, there the 



complex whole is strengthened by the property of 
the parts. 

[The word ak$dh conveys a whole consisting of three 
things. The plural number in the word shows that there are 
several things in that whole. These several things are con¬ 
nected in a certain order with the verbs. The order is the 
attribute of the several things included in the whole. 

The complex whole here is the first kind of whole men¬ 
tioned in the Vrtti which speaks about three kinds of wholes. 
In this first kind of whole, there are three distinct parts : As 
the FW/iputsit : tatrak$a ityavayavabhedanugata ekahsamudayah.'] 

Another way of looking at the same thing is now stated. 

466. When the complex whole is analysed into 
its parts, the meaning of the sentence is connected 
with each part. Or the mention of the accessories 
is split up into many because of the plural suffix in 
it, results in the splitting up of the sentence. 

[What is meant is that ak$a bhajyantam , bhujyantdm> 
divyantam becomes ak$o bhajyatdm , bhujyatam , divyatam.'] 

467. Either each word aksa can be used in the 
singular number or according to the original utter- 
ance, each word aksa can be separated in the 
plural number. 

[ What is meant is that each word aksa standing for one 
thing only can be used in the singular or plural number. 
Thus, it is shown that the splitting up of the original utterance 
ak$ah can take place in two ways. 

Normally, one would use a separate word for every idea or 
object which one wants to express : Pratyartham tabdanivetah , 
as it is generally put. Pratyartham is explained usually as 
artham artham prati. But it can also mean arthau arthau prati or 
arthan arthan prati. In the last two cases pratyartham Sabdanive - 
Jah would mean for every two ideas or objects or for many 

II. 468-471 


ideas or objects, one should use a word. It means that the 
-same word can express more than one idea or object. Those 
“who hold this view are abhedavadinah = those who hold that 
the word is the same even when it denotes many things. Accor¬ 
ding to them, the sentence ak$ali bkajyantdm , bhujyantdm divyan - 
tdm is not to be split up on the basis of the word ak$ah under¬ 
stood as including more than one word ak§a each being conne¬ 
cted with one of the verbs because, according to this view, 
there is only one word ak$a which has three meanings. That 
is possible only if it has three distinct powers. In other 
words, abhedavadinah believe in Saktitantra and not in Sabda- 
.t antra.'] 

468. According to those who believe in the 
identity (of the word) in the case of polysemic 
sentences, it is the power of the many-powered 
sentence which is split up (and not its form). 

[The other view is that in the sentence in question, 
the word akfah contains three separate words, all having the 
vform akfa. These three separate words are mentioned in a 
•condensed form but at the time of understanding the meaning, 
each of them should be taken out and connected with one of 
ithe verbs. This is the sabdatantra view. 

469. Or it is a case of mentioning in a condens¬ 
ed form (through tantra) two separate words. The 
inter connection between the words differs accord¬ 
ing to the hearer. 

470. What were two separate sentences having 
the same form have been used in a condensed form 
by the speaker for the benefit of the hearers. 

The views of the bhedavadin and the abhedavddin are now 
iput differently. 

471. Even if only one of the condensed senten¬ 
ces is meant by the speaker, the other (being of the 


vakyapadIyam of bhartrhari 

same form) comes to the mind. Without any 
intention on the part of the speaker, the other comes, 
to the mind as a power. 

[The first half of the verse gives the view of the bhedavadin- 
and the other half that of the abhedavadin. ] 

Another way of putting the view of the abhedavadin is. 


472. Sometimes, both the powers of the word 
are utilised at the same time just as fire is used 
sometimes both for its heat and for its light. 

It is now shown that the same thing happens in the Sdstra- 


473. The sentence, heard only once but having 
more than one meaning, eitherthrough repetition of 
the sentence or through more than one power, 
presents itself as more than one ( vibhagena ) either 
through lihga or through tantra. 

[According to the bhedavadins it presents itself as more 
than one through lihga and according to the abhedavddins it" 
presents itself as more than one through tantradharma, that is,, 
through its multiplicity of powers. ] 

An example is now given. 

474. In regard to the name Samprasdrana 5 . 
through two indications, the same sutra can be 
understood separately as presenting either the 
phoneme or the sentence meaning as the named. 

[The sutra: igyanah samprasdranam (P.1.1.45) teaches the- 
name samprasdrana. What is the named ? One view is that the 
phonemes i, u, r, l are the named. Another view is that the- 

II. 475-478 


sentence-meaning, namely, the coming of ik in place of yan is 
the named. There are indications for both in the sastra. ] 

475. Similarly, it has been shown in the 
Bhasya itself that conveying of more than one thing 
by mentioning the word only once has taken place 
in the satra dvirvacane aci (P. 1.1.59). 

[In P.1.1.59, even though the word dvirvacane is mentioned, 
only once, it has to be construed twice in slightly different 
meanings by following the principle of tantra which is similar 
to, but not identical with the grammatical process called ekasefa 
where a word, mentioned only once, may stand for the same 
word repeated. For example, in ratnau, the word rama, occurring 
only once, stands for two ramas and in ramdh, it stands for three 
or more ramas.~\ 

The following verses relate to the history of the Grammati¬ 
cal Tradition in India. 

476. After the Sahgraha declined when it came 
into the hands of Grammarians who were fond of 
abridgements and had acquired only little 

[The Saiigraha is mentioned in the Mahabhasya , I.p.6.1.12. 
We are told that there the question whether the word is eternal 
or only an effect is discussed as one of the main topics. 
Commenting on this, Bhartrhari says in his commentary on the 
Mahabhasya that there were 14 000 topics discussed in the 
Sangraha: Caturdasa sahasrani vastuni asmin samgrahagranthe 
(M.Bha. dipika, p. 21, 1. 4-5. B.O.R.I.Post Graduate and 
Research Department Series no. 8.) ] 

47 7. And when the Master Patanjali who knew 
all the traditions ( ilrtha-dar'sina) had incorporated 
into his Mahabhasya all the arguments and princi¬ 

478. It was found that those who were not 


vakyapadIyam of bhartrhari 

sufficiently equipped ( akrtabuddhlnam) could not 
arrive at proper decisions while studying that work 
at once bottomless because of its depth and clear 
because of its lucidity. 

[According to Punyaraja, Bhartrhari wants to say here 
that only his Guru Vasurata could really understand the 
Mahabhasya properly. ] 

479. When that sacred work which was an 
epitome of the Sahgraha was ruined by Baiji, 
Saubhava and Haryaksa who merely followed dry 

480. The Grammatical Tradition slipped away 
from the hands of the disciples of Patanjali and in 
time, the mere text of it survived in the South. 

481. Then Acarya Candra and other followers 
of the principles of the Bhasya obtained the true 
Tradition from the mountain and elaborated it into 
many branches. 

[Punyaraja says that by ‘mountain’, the Trikuta is meant. 
He also refers to the belief that there on that mountain the 
original Grammatical Tradition, composed by Ravana and 
engraved on stone, existed. Some Brahmaraksas brought it from 
there and gave it to Acaryas Candra, Vasurata and they under¬ 
stood the true nature of the Science of Grammar from it and 
imparted it to their disciples and incessantly elaborated it. In 
the end it became a discipline with many branches. ] 

482. After mastering those principles and 
cultivating his own discipline this collection of tra¬ 
ditions was composed by our Teacher. 

483. Here only the gist of a few of those tradi¬ 
tions is given. In the third Kanda , there will be 
full discussion. 

II. 484-485 


484. The intellect acquires critical acumen by 
familiarity with different traditions. How much 
does one really understand by merely following one’s 
own reasoning only ? 

485. The learning of those who have not sat at 
the feet of the earlier scholars but flit from one idea 
to another does not attain complete definiteness. 

- ooo 


Section 1. On the Universal 

1. Some analyse a sentence as consisting of two 
kinds of words, others of four kinds and others still of 
five kinds, just as a word is analysed into base, suffix 

[In the Vakyapadiya, the eight topics mentioned in 
Chapter I, 24-26, are dealt with, namely, the two kinds of 
meanings, the two kinds of words, the two kinds of relation 
and the two kinds of purpose. In Chapter I, the purposes 
and other matters were considered. In Chapter II, the sen¬ 
tence which is to be grammatically analysed ( cinvakhyeya) 
and its meaning which has a fixed character ( sthitaloksana) 
were determined. In Chapter III, words and their meanings, 
obtained by artificial division ( cipoddhclra ) are going to be 
considered. Words are obtained by artificially analysing 
sentences and a sentence may be analysed in different ways 
according to one’s point of view. Strictly speaking, a sentence 
is indivisible and it is such a sentence which is expressive 
(vdcaka ). The cognition of individual words and their 
meanings is really an illusion ( vibliramci ) and they are the 
result of an artificial analysis of sentences. This artificial 
analysis is a means of understanding and explaining the 
indivisible sentence, just as the division of a word into stem 
and suffix is a means of explaining the indivisible word. 
Sentences are infinite in number and no two sentences are 
alike. Their artificial parts seem to resemble one another 
and these parts are abstracted from the sentence as that is 
the only way of explaining the sentence. The individual 
word which is abstracted from the sentence on the basis of 
meaning is as unreal as the stem and suffix abstracted from 
the individual word on the basis of meaning by following 



the method of anvaya and vyatireka (agreement and differ¬ 
ence) mentioned by Katyayana in his Va. 9 on P. 1.2.45— 

Siddham tv anvyayavyatirekabhydm 

When the sentence is divided into individual words, the 
sentence-meaning is automatically so divided. The former 
depends upon the latter. That is why we cannot divide a 
word into so many phonemes, because these would have no 
meaning. The sentence-meaning is of a fixed character 
(sthitalaksana) . It is really indivisible, but we divide it 
artificially and we get two kinds of meanings out of it: action 
and accessory to action. That which expresses the former 
is called the verb (a khydta) and that' which expresses the 
latter is called noun (namari) . Thus we get two kinds of 
words. In a noun, there are two parts: the root and the 
suffix. Number and the like are the meaning of the suffix 
and even though, formally, it is the meaning of the suffix 
which is the predominant one in a noun, from the point of 
view of the meaning, that which comes from the word as a 
whole, namely, the individual ( dravyci ) qualified by the 
universal (jati), something which is in the nature of a thing 
(siddha) , is the predominant meaning, because it is this indi¬ 
vidual which is the substratum of number and the accessories 
(sadhana) . Particles ( nipdta) , Prepositions ( upctsarga) and 
Postpositions ( karmapravacanlya) come under nouns and 
verbs. Some particles express some peculiarity belonging to 
the thing ( siddha ), the meaning of nouns and, therefore, come 
under them. It makes no difference whether a word expres¬ 
ses a thing ( siddha ) directly or expresses a peculiarity in it. 
Some particles like svah are primarily expressive of things 
(sattvapradhana) . Thus they also come under nouns. Parti¬ 
cles like hiruk come under verbs, because they are primarily 
expressive of actions. It is not merely words ending in verbal 
affixes (tin) which are verbs. Any word which is primarily 
expressive of a process is a verb. That is why prepositions 
(upasarga) and postpositions (karmapravacarfiya) and parti¬ 
cles ( nipdta) are also verbs, because they also can express 
some peculiarity in a process. 

III. 1.1 


But, if one wants to emphasize their special feature, 
particles and prepositions are classed separately. They do 
not’ express a meaning directly, but express a peculiarity in 
the meanings expressed by nouns and verbs and so they are 
classed separately. Particles ( nipdta ) and prepositions 
(upasarga ) differ from one another also, because the former 
express a peculiarity found in things and processes whereas 
the latter can express a peculiarity in processes only. Post¬ 
positions ( karmapravacavdya ) on the other hand express a 
relation brought about by some particular action. Thus, they 
also denote a peculiarity in processes and come under pre¬ 
positions. In this way, we get four kinds of words, according 
to some. 

Others consider postpositions ( karmapravacaniya ) to be 
a fifth kind of word. They argue that they differ from upa- 
sargas because they refer to an action that is past and not to 
one that is present. The delimitation of a relation with refer¬ 
ence to some action or other, is their function. Relation is 
brought about by action or some service (upakfira) rendered 
through action. Sometimes the verb expressive of the action is 
actually heard and sometimes it is not. Where it is actually 
heard, the understanding of a particular relation is direct as 
in matuh smarati, matuh smrtam, sarpiso jdmte. Action can 
spontaneously and directly enter into relation with things, 
without another action coming in the middle. Where the verb 
is not used, there are two possibilities: the very nature of 
the things related brings the action to the mind and, without 
the help of the karmapravacaniya, we understand that the 
relation must have been preceded by a particular relation of 
action and accessory between two things, upagor apatyam, 
vrksasya sdkhd are examples. In upagor apatyam , there is 
the relation of father and offspring, it is brought about by 
the act of procreation. In vrksasya sdkhd , there is the rela¬ 
tion of whole and part, brought about by the action of stand¬ 
ing and supporting ( sthitikriydnimittah ). Sometimes, the 
relation has not the power of bringing any particular action 
to the mind. For instance, in the sentence rdynah purusah, 
the relation is that of master and servant ( svasvdmibhdva) 



and it may have been due to one of many actions: mainten¬ 
ance, abduction, purchase and begging. (Cf. M. Bha I. 
p. 463, 1. 12). From the expressioin ‘the King’s man’ 
(rajnah purusah) the particular action which was the cause 
of the relation of sva and svamin is not understood. Some¬ 
times, it is the presence of the post-position which tells us 
about the particular action which has brought about the 
relation in question. For example, in the sentence Sdkalyasya 
samhitam anu prdvarsat , there exists the relation of cause 
and effect between the samhita and rain and this relation is 
due to the action of reciting the samhita text. The post¬ 
position anu tells us that it is due to this action. (Cf. Vak. 

a. II, 109-201; 204-205). The word anu is found associated 
elsewhere with the action of hearing, as in anu-nisamya. But 
that is not a reason for considering it as directly expressive 
of the action here. Nor does it suggest action because a verb 
expressive of the act of reciting is not used here at all. Nor 
can anu suggest a verb here in the way in which vi suggests 
the gerund vimaya in jyrddesam vi-parilikhati . In Prddesam, 
there is a suffix (case-ending) expressive of an accessory to 
action; therefore, vi can suggest a word expressive of action. 
But in Sdkalyasya samhitam anu pnavarsat, there is a case¬ 
ending expressive of sesa in samhitam. Nor does anu directly 
express the relation of cause and effect, because that is done 
by the accusative case-ending which is specially taught 
instead of other case-endings. (P. 2.3.8.). The real function 
of anu is, therefore, to delimit the relation, to say that it is 
brought about by the particular action called ‘recitation’ 
(pathanam) . This is what is called Samhandhavaccheda , the 
delimitation of the relation. This is made clear in the verse 
Kriyayd dyotako nayam, samhandhasya na vdcakah etc. 
(Vak. a. II, 204). This is the function of anu . There is no 
use of attributing to anu something which is the effect of 
something else. 

Nor can it be maintained that the delimitation of the 
relation, referring it to a particular action, is the function 
of the sentence and that anu merely expresses the idea of 
posteriority. Those who maintain this rely on the principle 


III. 1.2 


that the meaning of the sentence is that which is understood 
over and above those of the individual words. But it must 
be remembered that ‘over and above’ does not mean going 
against the meaning of individual words. Sentence-meaning, 
even when it is an ‘extra’, must be based on those of individual 
words. It really consists of the relation of the process 01 
the thing primarily conveyed by a sentence with its acces¬ 
sories or its attributes. This delimitation of the relation can¬ 
not be considered to be the sentence-meaning, because it is not 
based on that of any individual word. It is, therefore, better 
to consider it as the meaning of the Karmapravacaniyci . 
About this delimitation of the relation by the post-position, 
there are two views: (1) that it is delimited in its own 
form ( svampe^dvacchidyate ); (2) that it is delimited as 

brought about by a particular action. According to Bhar- 
trhari it means that it is delimited by reference to the parti¬ 
cular circumstance (nimittavisesdvacchedah) . Thus in adhi 
brahmadatte pahcdldh , adhi indicates that the relation of ruler 
and ruled ( svasvdmibhdva ) is due to the act of protection 
which is the particular circumstance (nimittavisesa ). 

Thus the abstraction of words from a sentence on the 
basis of meaning is the only means of explaining a sentence.] 

The nature of the abstracted meanings is now explained 

2. When the meaning of a sentence is analysed 
into those of individual words, two word-meanings 
have been declared to be the eternal meaning of all 
words, either the universal (jdti) or the particulai 
( dravya ). 

[Once the meanings of words are abstracted from that 
of the sentence, the next question is: what is the natuie o 
these meanings? According to some, the universal (jati) is 
the meaning of all words. According to others, it is t e 
particular ( dravya ) which is the meaning of all words. A 
third view is that it is the particular as qualified by the 
universal. This third view is indicated by the dual numbei 



in the word paddrthau. If the third view is the correct one, 
then a word would convey both the universal and the parti¬ 
cular at’ the same time, because it is inconceivable that a 
word should convey one thing first and another thing after 
a little interval. When both are conveyed, one would be 
more important than the other in a particular context. The 
dual number in paddrthau might be taken to mean that the 
universal and the particular are equally important, but, in 
particular contexts, only one of them is conveyed. A word 
like gauh conveys the universal ‘cow-ness’, but, as a universal 
without a substratum is impossible, the particular is also 
necessarily understood. Similarly, the verb denotes the 
universal aspect of action, present in the different moments 
of action and causing the same cognition and the use of the 
same word. The universal of the accessory (baraka) con¬ 
veyed by a verb plays a subordinate part in the cogni¬ 
tion produced by a verb. The universal of action, conveyed 
by the verb, enters into relation with the universal of the 
accessory conveyed by the noun through the particular 
(vyakti ). In the particular inheres the universal which 
enters into relation with action through the power called 
accessory which also inheres in the particular. It will be 
explained in the section on the Accessory (S-adhanasamud- 
desa) that what is called ‘accessory’ is really a power or a 
capacity. The universal of an action like cooking is manifest¬ 
ed by the different' moments of it like putting the pot on the 
fire. Even though the universal of action is eternal, it be¬ 
comes a process ( sadhya ) through the particular. According 
to this view, prepositions ( upasarga ) also express the uni¬ 
versal, because they do no more than denote a peculiarity 
in the meaning of the verb which is a universal. A post¬ 
position is also based on the universal of a relation. Simi- 
larly, words like sukla express the universals of qualities 
(guna). So do proper names like Dittha. This is the view 
of Vajapyayana. (See M. Bha I. p. 242, 1. 10). 

According to Vyadi, the particular ( dravya ) is the mean¬ 
ing of all words, because it is the particular which is con¬ 
nected with action. What Vyadi calls dravya is often called 

in. 1.2 


Vyakti, the individual, also the correlative of universal. Thus, 
according to Vyadi, the particular or the individual becomes 
an integral part of the meaning of a sentence. It is to this 
that Vedic injunctions refer as declared in the VarUika. 

Codandsu ca tasyarambhat (Va 47 on P. 1.2.64). Even 
though the universal is not directly expressed by the word, 
it is also understood. In the verb also, it is the particular 
which is predominant, according to Vyadi. The meaning of 
a verb is the particular penetrated by action (vyapardvistam 
dravyam ). Words like sukla also denote the particular. 

Alternatively, one might understand by dravya what is 
going to be explained in the next section: that it is Brahman, 
conditioned by this or that limiting factor and that is the 
meaning of every word. 

Ordinarily, the word dravya is a synonym of vyakti , the 
individual. As the two word-meanings in question are 
declared to be those of all words ( sarvasabdanam ) they are 
also attributed to the parts like base and suffix which are 
abstracted from the individual word. The truth is that both 
are recognised meanings of words, even though one or the 
other may be emphasised in particular contexts. Whether 
the meaning of a word is jciti or dravya, in either case, it is 
something which is eternal ( nitya ). By nitya, it' is continuity 
(pravdhanityata) which is meant, as is made clear in the 

Tadapi' niiyam yasminstattvam na vihanyate 

(M. Bha I, p. 7, 1. 22). 

Thus, it has been shown that the universal and/or the parti¬ 
cular can be the meaning of a word. In this section, the 
considerations which favour the view that it' is the universal 
will be explained.] 

Now a doubt arises: If the universal is the meaning of 
every word, how can it enter into relation with action which 
is the meaning of the verb in the sentence? It is only the 
accessory ( sadhana ) which can be connected with action and 


the universal can never be the accessory. No doubt, its 
substratum can be the accessory; but, after all, it is the uni¬ 
versal which is directly conveyed by the word. So, when 
something having the universal prescribed in a Vedic injunc¬ 
tion is not available, we cannot use a substitute, because it 
would not have the prescribed universal. Thus Vedic injunc¬ 
tions cannot be carried out. This difficulty is removed as 

3. According to some, the universal is the means 
of conveying 'power’ because of association. If the 
‘acacia catechu’ etc. are powerless (i.e., not being 
available, are powerless to accomplish the action in 
question) a substitute having that power is taken. 

[According to the view of some theorists, the universal 
is the means ( upalaksana ) whereby power or capacity is 
conveyed, because the universal and power reside in the 
same substratum. Others, on the other hand, think that it 
is the universal which is primarily conveyed by the word 
and that, in some cases, it enters into direct relation with* 
action, while, in other cases, it does so through the medium 
of power. The word kesdmcit in the verse can also mean: 
‘in the case of some words’. Those words which express the 
accessories denote mainly capacity. In such words, it is the 
case-endings which stand for power or capacity and as the 
case-ending cannot be used by itself, the stem must neces¬ 
sarily be used and that expresses the universal which quali¬ 
fies the capacity conveyed by the case-ending. All this is 
true where the thing prescribed is available. Where it is 
not available, power or capacity becomes associated with the 
nearest universal. The Vedas enjoin actions which must be 
performed. Particular materials for the performance of these 
actions are also taught. If any material is not available, the 
action must still be performed with a substitute.] 

The author now points out how, even according to those 
who hold that’ the word primarily conveys the universal and 

not capacity through the universal the use of substitutes can 
be explained. 

4. (In the injunction ‘khadire badhnati’) the root 
‘bandh’ (to bind) denoting an action which results in 
loss of independence is taught like the act of killing 
(pramana) etc. Therefore, though the universal is the 
primary meaning, nothing that is devoid of ‘power’ is 

[Even according to those who maintain that the word 
primarily conveys the universal, the use of a substitute can 
be explained. In the injunction khadire badhnati, the action 
denoted by the verb is that of binding the animal to the 
sacrificial post, resulting in its loss of independence. That 
can take place only if a post which has the power to yield 
the desired result is used. Therefore, a substitute is allowed 
if the original article, the khadira wood, is not available.] 

The author now points out that there is an indication on 
the basis of which a substitute can be allowed. 

5. Even if the meaning of the root bandh is noth¬ 
ing more than contact (with the sacrificial post) the 
very fact that killing etc. are taught as the next step 
is an indication that an efficient thing is to be taken. 

[Even if it is maintained that the root bandh means noth¬ 
ing more than contact (with the sacrificial post and not loss 
of independence) the fact that killing etc. are taught as the 
next step is an indication that the animal must be well tied 
to the post. Therefore, something which is fit to be used 
as a post must be taken. Khadira has this fitness or capa¬ 
city, but if it is not available, something else which has this 
capacity must be substituted. Thus, on the view that a 
word denotes the universal, the use of a substitute becomes 

v-m. 2 



justifiable in three ways: (1) on the basis of the meaning 
of the word (; paddrtha ) as shown in verse 3 where it was 
said that the meaning of the word, the universal, stands for 
something which has the required power, (2) on the basis 
of sentence-meaning ( vdkydrtha ) as shown in verse 4 where 
it was said that the word must convey something which has 
the capacity to help in the accomplishment of the main action 
conveyed by the sentence and (3) on the basis of context 
(prakaraya ) as shown in the present verse which says that 
a substitute must be allowed if one takes into consideration 
the other acts like killing which are taught in the same 

Here Helaraja mentions two further arguments in order 
to justify the use of a substitute. The first one is called: 
asambhavaniyaviatyagah. An injunction like khcidire badhndti 
must not be understood as an asavibhavaniyama, that is, a 
restrictive injunction ( niyamci ) making alternatives impos¬ 
sible ( asambhava ). The second argument is called niyama - 
matratyciga. The injunction in question is so interpreted 
that its positive aspect is retained; that is, one understands 
from it that the animal should be tied to a post. One rejects 
the negative or restrictive aspect of it. In other words, 
one rejects the restriction that the post should necessarily 
be made of khadira wood. By adopting these two argu¬ 
ments, one avoids going against scriptural injunction, 
because one follows the positive aspect of it, eveh. if one 
rejects its negative or restrictive aspect.] 

6. All words first express their own universal 
which is then thought of as being superimposed on the 
universals of the meanings. 

[When we hear a word, it is its own form which we 
understand first. And we do so, no matter who utters it. 
The form remains the same in all utterances. This form 
is, therefore, a universal (j dti) and it is this which we under¬ 
stand invariably. This invariable cognition of the uni- 

III. 1.7-8 


versal of the word is what is meant by the word ‘first’ 
( prathamam ). See Vak I. 66. After this is understood, or 
while it is being understood, the meaning which is also a 
universal, is understood. The two appear to be one. The 
form of the word is understood by us as having been super¬ 
imposed on the meaning. We take the two as one. To under¬ 
stand the one as being superimposed on the other implies 
that we consider the two to be different. We do, but this is 
only a fiction ( kalpand ) on our part, because the meaning 
(the object), according to Grammar, is only an unreal mani¬ 
festation (vivartci) of the word. Therefore, the two are not 
really different from each other. This superimposition of the 
word on the meaning is the basis of convention and also of 
the relation of expression and expressed ( vcicyavdcakablidva) 
between the two. This universal of the word is progressively 
clearly revealed by each sound of the word, just as each 
moment of a movement like the lifting of the arm reveals it 
fully and progressively more clearly, or just as a passage of 
a text, not fully learnt by heart at the first reading is gradually 
memorised by repeated reading or just as the genuineness of 
a precious stone, not fully clear at the first gaze, becomes so 

The author now says how the universal of the word, 
existing in the word, can convey the universal of the object 
as identical with itself. 

7-8. Just as ‘redness’, residing in the quality red 
is attributed to the substance ‘lac’ and then, on ac¬ 
count of its intimate union with the intimately united, 
it is perceived in clothes etc., in the same way, on ac¬ 
count of the intimate connection between word and 
meaning, the universal residing in the word performs 
the function of universal for the universals of objects 

[It was declared in the previous verse that the universal 
of the word is superimposed on that of the meaning and that 



the latter appears as one with the former. Here a doubt might 
arise: the universal of the word is inherent in the word and 
not in the meaning; how can it then denote the meaning 
through the relation of identity? This doubt is sought to be 
removed by means of an example. Redness is inherent in a 
particular case of red, in what is red. It is also attributed to 
the substance which is red. Due to contact with the sub¬ 
stance which is red, redness is attributed to another object 
also like cloth when we speak of it as being red. In other 
words, due to the contact of the cloth with a substance which 
is red, the cloth is also said to be red. Similarly, there is the 
eternal relation between the word and its meaning, a relation 
which is grasped only at the time of the learning of the con¬ 
vention. This relation is evident from the fact that both about 
the word gauh and the object gauh, we can say: ayavri gauh. 
Through this relation, what belongs to the word is super¬ 
imposed on the meaning. The universal of the word is super¬ 
imposed on the universal of the meaning. The latter becomes 
the expressed meaning ( vdcya ) of the former. Not only that. 
The universal of the word becomes a kind of universal among 
the different universals, though, according to the Vaisesikas, 
there cannot be a universal of the universals. But the uni¬ 
versal of the word plays the part of a universal among the 
universals of the meanings. It performs the two functions of 
such a universal. The two functions are: persistence of the 
same word ( sabdanuvrtti ) and persistence of the same cogni¬ 
tion (pratyaydnuvrtti) . How this happens is explained in 
verses 9 and 10. 

This universal of the word exists in the word considered 
as a unity and not in the different phonemes, because it is this 
unity which is expressive (vacalca). This unity is called 
vyaktisphota and the universal, jati, inheres in it. It is called 
jatisphota. Thus sphota is of two kinds. The particular rela¬ 
tion through which vyaktisphota expresses the meaning is fit¬ 
ness ( yogyatd ) which is eternal and not the work of man 
(apauruseya) . When a meaning is understood from a word, 
it appears as one with it. With all that, the form of the mean¬ 
ing is not completely obliterated. When an object is revealed 

III. 1.9 


through light, it appears as enveloped in light, but with all 
that, its own form appears distinctly. Light and the word 
reveal other objects by superimposing their own form on them. 
This is the eternal character of words, not due to human 
agency. This can happen only in the case of one who knows 
a language. In the case of one who like a child does not' know 
it yet, there cannot be superimposition of the word on the 
meaning, because he does not know the word yet. The uni¬ 
versal of the object appears to him as distinct.] 

How it performs the function of a universal is now 

9. When one has to form the ‘ekasesa’ of words 
denoting universals, the universal of words is the 
universal of the universals of objects. In the formation 
sabdajatayah (universals of words) the universal of 
the word (is the cause of the ekasesa ) when the uni¬ 
versals of the words have to be expressed. 

[When we apply the same word, say ‘cow’ or ‘tree’ to a 
large number of objects, it is because we see some common 
characteristic in ail of them called jdti or universal. When we 
apply the word 'jdti to a large number of universals, the 
natural conclusion would be that, in all of them, there is a 
common characteristic, another universal above the original 
universals. But this is impossible according to the vaisesika 
principle: nihsdmanydni sdmdnydni. And yet we do apply the 
word jdti to a large number of universals. How to explain this? 
It is explained in this way: the universal for which the word 
jdti stands does the work of such a super-universal which really 
does not exist. Similarly, if each word stands for its universal, 
there would be as many word-universals as there are words 
and when all these universals existing in the different words 
are referred to by the expression sabdajdtaya imali , what really 
happens is that the universal existing in the word sabdajdti 
itself does the work of such a super-universal.] 



10. The universal of words, residing in words and 
differing from the words themselves, includes also the 
universal existing in the word sabdajati. 

[The universal which exists in words and which is different 
from the words themselves is so comprehensive that it includes 
the universal which exists in the word sabdajati itself. When 
we say: sabdajati, it does not denote a universal which is over 
and above such universals as gosabdatva, asvasabdatva etc. 
Because, one does not believe in a universal over a universal. 
So the universal which exists in the word sabdajati itself does 
the work of such a universal. The universal which exists in 
the word sabdajati is in the same category as the universal 
which exists in the word gauli or asvab. Atikraviya vartate 
has been translated as ‘includes’, following Helaraja.] 

The author now tries to show that' all words denote the 
universal without resorting to the concept of ‘adhyasa’. 

11. (Some) words may denote the universals of 
objects, but all denote the universal, since the mean¬ 
ings of words are determined by the functions ( vyd- 
para)' of words (i.e., they are determined by what the 
words actually convey)'. 

[So far, the author has established the doctrine that all 
words denote the universal with the help of the grammarian’s 
idea of adhyasa, the idea that the universal of the meaning 
a ppears as one with that of the word, or rather that the 
universal of the word is superimposed upon that of the mean¬ 
ing. Even without recourse to the idea of adhyasa, it is 
possible to show that all words denote the universal. If one 
does not accept adhyasa, the alternative is to say that words 
denote the universals of the object, pure and simple. Even 
then, one will have to maintain that all words do so. By ‘all’, 
hat is meant is: not only words like ganh, asvah, etc., but 
also words like jati. Here one has to meet the Vaisesika ob¬ 
jection that if the word jdti also stands for a universal, it' must 

be a universal existing in other universals. And that is not 
possible. If there is universal in universals, where would one 
stop? But grammarians have a different point of view. Their 
chief concern is to find out the nature of meanings conveyed 
by words. What they find is that in all universals as conveyed 
by words, there is a common point or characteristic which can 
be looked upon as another universal and can be called by the 
name jdti. The existence of the first universal was postulated 
because a common characteristic was experienced in the indi¬ 
viduals. Even invisible things are assumed to exist if they 
somehow produce a cognition in regard to themselves. What 
to say of things conveyed by words? Grammarians go by what 
the words convey. They are not really concerned with things 
as they really are, but with things as conveyed by words. If 
a quality is conveyed by words as a common characteristic, it 
becomes a universal for them. The word syama is considered 
to be expressive of a quality but Panini gives to .its meaning 
the name of samanya in the compound word sastrlsyamd ac¬ 
cording to his sutra: upamdnani samanyavacanaih . (P. 2.1.55). 
The same is the case with action. In such matters, gramma¬ 
rians are more anxious to follow worldly usage than the views 
and conventions of other systems of thought. To them, artha 
means what words convey and all words convey the universal, 
because that is a matter of fact. In reality, the universal which 
a word conveys may or may not exist. But the word in ques¬ 
tion, through its function called abhidha , does convey it. So 
that is the meaning of the word.] 

The author now makes clear the function called abhidhd 
of words. 

12. In the view that the universal is the meaning 
of words, the universal or even the particular becomes 
the meaning of words in the same way as (the well- 
known) universals; therefore, they (the words) denote 
the universal. 

[The peculiar features of an object, not found in any other 
object in the world, cannot be expressed by words at all. At 


the time of learning the convention about words, it is with the 
common characteristics of objects that we connect the words. 
Thus the words become expressive of them. The universals 
may really differ from one another, but' words cannot express 
them. Even proper names denote universals. The name of 
a person stands for that unchanging recognisable element 
which persists in all the changes which he undergoes.] 

The author now states that even the other view, namely, 
that words denote substance ( dravya ) depends on the function 
of words. 

13. In the view that substance is the meaning of 
words, the meaning of all words can have the characte¬ 
ristic of substance. As the meanings of words are sus¬ 
ceptible of assuming the properties of substance, the 
latter is said to be the meaning of all words. 

[The other view about the meaning of words is that it is 
dravya which can be rendered as substance (as distinguished 
from attribute) or individual (as distinguished from the uni¬ 
versal). This is the view of Vyadi as stated in Va. 45 on 
P. 1.2.64. According to this view, all words denote dravya, 
some primarily and directly, others secondarily and indirectly. 
Just as, in the other view, some words denote the universal 
primarily and directly, while others do so secondarily and 
indirectly. According to this view, whatever may be the 
reality, a thing as conveyed by words is seen to possess the 
characteristics of dravya which are: the capacity of being 
referred to as ‘this’ or ‘that’, a certain completeness and inde¬ 
pendence, the fact of having gender and number. A quality 
like ‘whiteness’ may or may not in reality possess these attri¬ 
butes, but as presented by a word, it can have these properties 
or rather these properties are attributed to it.] 

So far, it has been shown that all words have the same 
meaning (the universal or substance) on the basis of word- 
function ( sabda-vydpdra ). The author now proceeds to show 

in. i.i4 


that, according to the Vaisesikas also, i.e. on the basis of the 
common characteristic of objects, all words denote the 

14. In all universal, the fact of being a universal 
consists in being found in all (the individuals which 
belong to the class) while in all particulars ( visesa ), 
.the universal consists in this that they distinguish (the 
things in which they reside from others). 

[So far it has been maintained that it is in the nature of 
a word to convey the universal, whether that universal really 
exists or not. A word conveys it through its function, its power 
to convey meanings. But even by following the Vaisesika line 
of reasoning, one can show that the universal is the meaning 
of words. What is, after all, a universal? It is something 
which exists in all its substrata, as a result of which all of 
them produce a uniform cognition and are called by the same 
name. Another characteristic of a universal is that it pervades 
the whole of its substratum and not merely a part of it. A 
universal like gotvci has these characteristics. But there are 
many universals and it is not unreasonable to believe that 
there is another universal existing in all of them and sharing 
the characteristics of a universal in general. The word visesa 
also denotes a common characteristic of the same kind. There 
are as many visescis as there are eternal things ( nityadravya ) 
and they all have this common characteristic, namely, that 
each one exists in one eternal thing and not in any other. 
This is also thus a common property, something like a uni¬ 
versal. Words like ablidva also denote a universal of the same 

What Helaraja wants to point out is that all the visesas 
produce a uniform kind of cognition and are the cause of the 
same name (visesa) being applied to them.] 

The author now explains how words like ak&sa, kcila and 
difc, which stand for things which are one and eternal, denote 
the universal. 

v-m. 3 


15. Once parts are postulated on the basis of the 
properties of the different objects (with which space 
(akasa) is) in contact, the universal of space also is 
found in these parts. 

[Words like akdsa, kdla, dik, stand for that common 
feature which is present in all the contingent divisions which 
we make in these things. For instance, akdsa is one, but we 
artificially divide it into localities according to the different 
objects which exist in space. The space occupied by a jar is 
dkdsa just as the space occupied by a piece of cloth is akdsa. 
In regard to all these divisions, we have the cognition that they 
are dkdsa and that is what the word stands for. The same 
thing can be said about time, direction, soul and the relation 
called samavdya. Soul is one, but it can be looked upon as 
different in each person with whom it is associated. We can 
say also that there are as many samavdya-s as there are pairs 
of things united by samavdya; but everywhere there is 
ihabuddhih: which is the connecting link.] 

It is also possible to look upon these divisions of dkd-sa 
as real ones and not merely as contingent ones. 

16. Just as the connected (potsherds etc.) are 
parts of the jar and the like which (as such) are with¬ 
out divisions, in the same way, the connected objects 
are the parts of akasa which is really without any 

[In the previous verse, it was declared that words like 
dkdsa denote a kind of universal on the basis of the contingent 
divisions of such things as akasa. But, perhaps, one can speak 
of these divisions as real and not contingent. An object like 
a jar, a whole, is, in itself, partless. Its parts, the potsherds, 
are its causes and connected with it by the relation of inherence 
(samavdya) . Similarly, the various objects of the world 
which are in contact with akasa and which delimit it are the 

III. 1.17-18 


real divisions of aktisa. They become the desa of dkasa which, 
being their background, become their desa too. Ak&sa and 
the objects become desa to each other.] 

The author now states where the divisions (desa) are 
primary and where they are secondary, as, unless they are 
primary somewhere, we cannot speak of them as being 
secondary elsewhere. 

17-18. The idea of difference persists in the case 
of objects joined together by the relation called con¬ 
junction ( samyoga ) whereas in the case of objects 
joined together by inherence ( samciv&ya ) the percep¬ 
tion of difference disappears. 

Therefore, the divisions consisting in objects joined 
together by conjunction are secondary whereas the 
divisions consisting in objects joined (to the whole) 
by inherence are primary because (the wholes) so 
united are not really different from the divisions. 

[When we have the idea of two things being in contact, 
it is based on a difference between the two things. Even 
when we see the contact, we continue to see them as differ¬ 
ent from each other. The two things which come into con¬ 
tact make a kind of whole, but the two things are not essen¬ 
tial parts of that whole, because we continue to perceive them 
as apart from the whole. But when two things are related 
to each other by inherence, the position is different. When 
we see a sphere, we do not see its two halves as distinct 
from the sphere. They form the very body of that sphere. 
They are, therefore, real parts of the whole called sphere. 
The division is a primary one and not a secondary one. But 
even within the field of things related by Samyoga or con¬ 
tact, one must make a distinction. When an all-pervasive 
thing like dkasa comes into contact with things, it is an 
invariable contact. The contact between the clothes and the 



body is not an invariable contact. In the former case, we 
do divide the all-pervasive substance on the basis of the 
objects with which it is in contact]. 

The author now explains the universal according to the 

19. Some consider the universal to be merely 
something which figures as a common characteristic in 
our mind and the particular ( dravya ) to be that which 
figures as differentia. 

[According to the Vijnanavadins, this attempt' to show 
that words like Skasa also denote the universal is futile, 
because, according to them, there is no such thing as the 
universal at all. They do not believe in the reality of the 
external world. They only believe in the different states 
of consciousness and, in them, some things figure as common 
properties while others appear as distinguishing features. A 
word denotes only this thing which figures in the conscious¬ 
ness and this is what is called jati. In such matters, the 
grammarians are governed by two principles: (1) Grammar 
is common to all disciplines, (2) for Grammarians, meaning 
( artha ) is just what the word conveys. The second princi¬ 
ple enables them to explain the universal according to the 
Vijnanavadins. According to the latter, the universal is 
something which occurs in the mind when a word is heard. 
In other words, it is sabdcirtha and for Grammarians also, 
artha means s abdartha.'] 

The author now points out the need for' postulating the 
existence of the universal. 

20. To say that things are different or that they 
are identical presupposes some external basis (j^cno- 
pddhih ). This diversity in objects arises only when 
they are united in some way. 

III. 1.21-24 


21. Neither identity nor difference, neither exis¬ 
tence nor non-existence of objects is possible if they 
are not linked with one another. 

[The above two verses are meant' to be an answer to 
one who might say that things differ from one another only 
because of time, place and circumstances. If such differ¬ 
ences are ignored, all things would be the same. W y tnen 
postulate an universal? It is pointed out here that unless 
it is postulated, all vyavaliara or verbal transaction, woul e 
impossible. Vyavahara depends upon some basis (uyudhi) 
or other]. 

The author now puts forward the advaita view on this 
subject, namely, that it is the powers of the One which are 
the basis. 

22. The ultimate view is that it is the One which 
has all the powers. To postulate difference between 
the objects themselves is unnecessary. 

23. Therefore, substance etc. are only powers (of 
the Supreme), made known by their different func¬ 
tions. United, they help man to reach his goal and not 

24. Just as the collocation of the senses etc. is not 
an entity over and above the things composing it, in 
the same way, the connection between the different 
powers of the Supreme is not a separate entity. 

[All the previous discussion arose in connection with 
the question: how do words like ak&sa denote the universal? 
The answer was based upon contingent divisions of dkasa. 
Contingent’ divisions are only artificial divisions which really 
do not exist. But when it comes to the other things of the 
world, they were assumed to be real. In verse 22, the author 
states that ? according to Advaita, Brahman is the only reality. 


Everything else, that is to say, all distinctions such as the 
universal, the particular and so on are unreal. They are all 
the products of the powers of Brahman. The seven catego¬ 
ries of the Vaisesikas are only the powers of Brahman. These 
powers are inferred from their effects. Though the various 
objects of the world, produced by the powers of Brahman 
are distinct from one another, still they join together in order 
to do their work. But one cannot, because of that, maintain 
that it is their connection and not the things themselves which 
are effective nor that the connection which is eternal, is an 
entity different from Brahman. That would go against 
monism. The connection is not over and above the things con¬ 
nected. Just as in perception, the senses, the mind and the 
object must come together and this coming together, this 
collocation (sdmagn) is not an entity over and above them 
so is the case here. Thus, there is no damage to monism]. 

Remark: Sambandhi-sambandha-samsarge’pi. Samban- 
dhin =the Supreme, sambandlia =the powers of things, sam- 
sarga =the connection of the powers. 

The author now says how the universal which is eternal 
helps the effect. 

25. Nothing is produced in which its universal 
does not exist. Hie universal sets the causes (of the 
particular) in motion for its own manifestation. 

26. The universal, after having obtained a foot¬ 
ing among the eternal and non-eternal causes, manifest 
themselves again and again in some effects. 

[The universal requires the particular or the individual 
for its manifestation. It, therefore, prompts, as it were, the 
causes of the particular, so that the latter may come into 
existence. It becomes a kind of contributory cause in the 
creation of the particular. The word used here is prayojikd. 
In Vaisesika terminology, it would be a nimittakdrana. The 
universal exists in the material cause of an object, whether 
eternal or otherwise and makes its appearance when the parti- 


III. 1.27-28 

cular is produced. The effect is supposed to exist in the 
cause in the form of the universal which is manifested in 
the particular which is produced. Thus, the universal helps 
in the production of the particular which is the effect]. 

As the universal exists even before the individual is 
produced, it is eternal. How it helps is now stated. 

27. The universal is the cause even of that object 
which is newly produced. In order that its substratum 
might be produced, it prompts the action to come to be. 

[In a sentence like sa Katam Karoti (he makes a mat), 
‘mat’ is the object of the action of making. In other*words, 
it is a scidhcina , an accessory of the action of making. Before 
it is made, it does not exist. How can it then become a 
scidhcina ? To this objection, the answer is that its universal 
was already present in its causes and it is that which helps 
in its production. The universal aspect of it is the sadhana 
and the individual ( vycikti ) aspect of it is the effect ( Kcirya ). 
After all, the universal and the individual are not two totally 
different things like a cow and a horse. They are indis¬ 
solubly mixed up. They are two aspects of the same thing 
which gets two names accordingly. Why the universal is 
called prciyojikd is this: Before a man makes a mat, he has 
the universal of it in his mind and then proceeds to assemble 
the material. But for that, he would not act. It is the 
universal which prompts him]. 

The author now states that, according to the view that 
the word denotes the universal, it is not the individual which 
is the accessory of an action ( sddhana ), but the universal. 

28. A universal such as brahmana-hood is the 
accessory in the case of a command or a prohibition. 
This universal is qualified by the one belonging to the 
number of the individual. 

[It is the universal which is the object of all the actions, 
commands or prohibitions, which are found in the Vedas . In. 



carrying them out, it is the universal which is the accessory, 
that is, the means ( sddhana ). Not only that. The universal 
of the number expressed by the suffix qualifies this universal] . 

As the universal is never found apart from the indivi¬ 
dual, would it be right to maintain that the word expresses the 
former only? This doubt is now answered. 

29. Just as a face, reflected in water etc. is only 
called as such (that is, a face without any connection 
with water etc.), in the same way, a word expresses 
only the universal manifested by the individuals. 

[In reality, we never see the universal apart from the 
individual in which it inheres. But in spite of that, a word 
conveys only the universal through its expressive power 
( abhidha ). We use the word mukha for the face which is 
reflected in a mirror and not also for the mirror which re¬ 
flects it. Similarly, a word conveys by abhidha, on the basis 
of convention, the pure universal and not also the individual 
which manifests it'. In perception} however, the two are 
mixed up. A word, on the other hand, conveys a meaning 
according to convention. 

See Vak. I. 100 and the Vrtti thereon]. 

The author now states how, if the individuals are not 
conveyed by the expressive power of words, they can dis¬ 
tinguish the universals. 

30. Just as the difference in the organ of sense, 
though unperceived because of the non-perception of 
the organ still leads to an awareness of the difference 
in the knowledge (produced by the senses) of their 
respective objects, 

31. in the same way, some individuals, though 
themselves unperceived, still become the cause of the 
difference in the knowledge of the universals. 

III. 1.32-33 


[The individuals, though not conveyed by words, still 
serve to qualify the knowledge of the universal which arises 
from words, just as, in perceptual knowledge, the senses, 
though not themselves perceived, still serve to distinguish one 
kind of sense perception from another]. 

So far, the view that the word conveys the universal only 
by its expressive power has been explained. There are two 
other views: (1) that it conveys the individual as quali¬ 
fied by the universal, (2) that it conveys the individual. 

The author now states that the distinction between the 
universal and the individual can be explained on the basis 
of the view that everything is the result of the different 
powers of the One Reality. 

32. Of the real and unreal elements which are 
found in every object, the real element is the universal, 
whereas the unreal one is the individual. 

33. It is Being which, being differentiated accord¬ 
ing to the object in which it is present, is called the 
universal. All words are based on that. 

[That element in objects which is peculiar to them, which 
comes and goes, is a transformation of nescience (avidyd ). 
It is that which is called vyakti. It is not real, because it 
is not permanent and not found in all objects. Only that is 
real which is permanent and which constitutes the pervading 
essence of an object. That is what the monists call the uni¬ 
versal (jati ). In such gold ornaments as rucaka, svastika , 
kuiidala etc., the particular shape of the ornament is the 
impermanent thing. These shapes are mutually exclusive, 
whereas gold is the persisting essence of all. That is real 
or satya. But it has only a relative reality. Gold is fire 
(iejas) which is more pervasive and, therefore, the persist¬ 
ing element compared to goldness. But even the reality of 
tejcis is relative. If we go on seeking wider and wider reali¬ 
ties, we will come to Brahman which is the only Reality 
V-m. 4 


vAkyapadiyam of bhartrharj 

which is everywhere and in everything. That is the highest 
universal. So it is called maha-satta, the great Being. It is 
this mah&satta. which is found in all objects in the form of 
their universals, as ‘cowness’ in a cow, ‘horseness’ in a horse 
and so on. What is called gotva is nothing but gosattd Being 
as it exists in a cow. All words thus ultimately express this 
great Being and it is only through the limiting factor with 
which it is associated in each object that it appears as golva, 
asvatva and so on]. 

34. That is the meaning of the stem and of the 
root; it is eternal, it is the great Soul; it is the meaning 
of the suffixes -tva and -tal. 

[In all things, there is the comprehensive universal called 
‘Being’. Even negative entities have it because they also 
figure in our mind. Hence every noun-stem denotes this, 
whether it is the name of a positive or negative entity. Roots 
also do the same. They denote ‘Being’ as existing in the 
different individual actions which depend upon the accesso¬ 
ries ( sadhana ) for their accomplishment. Thus, it is this 
Being which is presented as finished or accomplished ( siddlia ) 
by nouns and as a process ( su-dhya ) by roots and there is no 
third possibility. It is not merely the stem and the root' 
which thus denote Being, but also the suffixes. They denote 
Being as associated with such limiting factors as number, 
accessory etc. Not merely according to the monists, but 
according to the followers of Sankhya also, all words denote 
Being which is the same as what they call ‘Mahat’ or Buddhi]. 

Remark: For the Sankhya view, see Vyasabhasya on 

Y. Dar. II. 19. 

This idea is now further explained— 

35. It is this universal, (namely, Being) which 
is called ‘action’ when it assumes sequence in the 
details. When the form having sequence is destroyed, 
it is just called substance ( sattvcwn ). 

III. 1.35 


[In all things, there is the comprehensive universal called 
‘Being’. What is called ‘action’ and what is called ‘substance 
(sattva) —both these are only transformations of Being 
(sattd ). A verb expresses this sattd as a process, something 
which has parts arranged in a sequence and which depends 
upon accessories endowed with movement foi its accomp is 1 
ment. Every verb expresses a process of this kind. Even 
though these processes differ from one another, they are all 
processes. The fact of being a process is their common fea¬ 
ture. This is the universal of action which is a vivaria, an 
unreal manifestation of sattd or Being. This is the meaning 
of all roots. When a process of this type is not understooc 
from a root, when the inner sequence which is the essence 
of a process is not understood, what we understand is called 
sattva, substance or thing, which is also a transformation o 
Being. This is also called dravya. Action is sddhya an 
sattv°a is siddha, and they are the two unreal manifestations 
(vivarta) of Brahman. To say that all words convey Being 
is equal to saying that all words convey the universal (jott) 
because Being is the greatest of all the universal and it is 
identical with Brahman. Thus according to the view that 
all words convey the universal (the jdtipaksa), they real y 
convey Brahman. According to the other view also, namely, 
that all words denote dravya, they denote the same thing, 
namely, Brahman, because it is going to be explained t la 
what is called dravya is also Brahman. Jdti and di avya are 
only two ways of understanding Brahman. When it is 
thought of as the persisting feature in everything, it is called 
jdti. When it is thought of as a finished thing, it is called 
dravya. Thus both the views really stand for the same view, 
namely, that' all words denote Brahman. 

All this has been said according to Brahmadarsana. But 
Vydkarona, as a sdstra, is not particularly attached to any 
system of philosophy. It is sarvapdrsada, it is common to 
all schools. Therefore, points of Grammar are sometimes 
sought to be explained by reference to other systems 


vakyapadiyam of bhartrhari 

The author now states that, according to the author of 
the Nirukta also, all actions are unreal manifestations of 

36. It is this universal which assumes six states 
in the transformations of Being. On account of its 
powers,- it appears so in a sequence. 

[That all actions are only unreal manifestations of Being 
(sattd) is supported by Yaska also. He quotes with appro¬ 
val the view of Varsyayani who said that bhava or Being 
undergoes six transformations: it is born, it exists, it changes^ 
it increases, it decreases and it' is destroyed. A thing is first 
born, then it is said to exist, what exists necessarily undergoes 
change, change means growth and then decay and, finally 
comes destruction. (See Ni. 1.2.9.) These so called transfor¬ 
mations ( vikara ) are not real. Here viktira only means 
vivarta, an unreal manifestation. Followers of Agama main¬ 
tain that Brahman who is omnipotent has a power called 
‘Nescience’ on account' of which the one Brahman is presented 
as many. Because what is manifested has multiplier tv, the 
power which is One, is spoken of as being many]. 

The author now explains how the manifestations appear 
in a sequence. 

37. Sequence also belongs to the very essence of 
this Being. In it, Time is seen, as though divided into 
forms like priority and posteriority. 

[What is called Time is a power of Brahman. Sequence 
is the very essence of Being and it is also the very essence of 
Time. Sequence is really not different from what has sequ¬ 
ence. Time is not an entity quite apart from the things which 
appear in a sequence. Time is just a power of Brahman, 
as a result of which actions appear with sequence within 

in. 1.38-39 29 

The author now explains how destruction is also essen¬ 
tially “Being”. 

38. If one accepts the eclipse of objects, that is 
their non-existence. When this eclipse assumes inner 
sequence, one gets the cognition that the object is 

[The question now arises: how can birth and destruction 
which are not positive things be transformations of Being? 
The answer, as far as destruction is concerned is this: what 
is called destruction is nothing more than hiding (tirobh&va ), 
not being visible. When a thing exists in its causes as a 
mere potentiality, it is called destruction. This condition 
exists before a thing is produced and when it is destroyed. 
When, due to the activity of the accessories, the process of 
hiding becomes perceptible and is conveyed by a verb, in 
its first stage it is expressed by the verb apaksiyate —it dec¬ 
lines, and, in its last stage, by the verb vinasyati—it is des¬ 
troyed. Thus, out of the six transformations of Being; two 
relate to the process of hiding or disappearing. Hiding or 
existing in the causes in the form of a potentiality is a form 
of ‘Being’]. 

The author now points out that ‘birth’ is also a trans¬ 
formation of ‘Being’. 

39. It is Being which, fallen from its previous 
state and not having yet attained the later state, is 
called ‘birth’ on the basis of the different moments 
coming between the two. 

[To say that something is being born amounts to this: 
that it has moved from the previous condition of existing as 
a mere potentiality in its causes, but has not yet reached a 
condition where it can be said to exist. Birth is not the 
coming into being of something which did not exist at all 
before. It stands for that intermediate state between existing 


vakyapadiyam of bhartrhari 

in the causes as a potentiality and existing as an actuality, 
a state conceived of as a process. In the list of the six trans¬ 
formations of Being, each represents the coming into being 
of some new feature. In that sense, each stands for birth]. 

The author now points out which associated things bring 
about multiplicity in ‘Being’. 


40. It is its substratum or its (fictitious) parts or 
things appearing to be different from it or its own 
powers which are the causes of differentiation in Being. 

[The substratum causes variety or multiplicity. Just as 
the same face, reflected in oil or water or mirror, appears 
to vary, similarly, the same ‘Being’, manifested by a cow or 
a horse, appears to be different. Or, one can say that the 
universals like ‘cowness’ or ‘horseness’ etc. which are nothing 
else than artificial divisions of Being, make the one appear 
to be many. The whole which is one appears to have multi¬ 
plicity through its own parts. Sometimes multiplicity is due 
to things like snace, time and the senses. We speak of things 
thus: ‘this is here now’, ‘that was there them’ and so on, show¬ 
ing how time and space bring about difference. To one who 
is suffering from jaundice, white appears to be yellow. The 
ultimate truth is this: Brahman which is omnipotent appears 
as many for the worldly experiencer]. 

After having pointed out that, according to the views 
of others and of the Grammarians, the universal is the mean¬ 
ing of words, the author now explains its properties. 

41. The universal does not depend upon any 
particular shape for its manifestation in Earth etc. Nor 
does it become destroyed with its substratum, though 
the latter is not eternal. 

[Universals like earth-ness, water-ness do not require 
any particular shape or disposition of parts for their manj- 

III. 1 . 42-43 51 

festation. A jar appears as earth just as much as a mountain 
does. But such universal as cowness (gotva) require a 
particular shape for their manifestation. Secondly, a uni¬ 
versal like a quality exists in its substratum by the relation 
of inherence (samav&ya), but if the substratum is destroyed, 
the quality is also destroyed, but not the universal. A uni¬ 
versal depends upon a substratum only for its manifestation 
and cognition and not' for its very existence]. 

Where would the universals, inhere when, at the Great 
Dissolution, even eternal things are destroyed? To this, the 
author replies as follows: 

42. Others declare that the substratum of all 
universals cannot be destroyed, holding tne view, that, 
at the Great Dissolution, everything is not destroyed at 
the same time. 

[The universals require a substratum for their manifes¬ 
tation. But, at the time of the Great Dissolution, all effects 
and transitory things which form the substrata of the uni¬ 
versals disappear. What happens then to the universals? 
This question is answered by some by saying that such a thing 
as Great Dissolution does not take place at all. So all uni¬ 
versals always have some substrata or other. Otheis say 
that all things are not simultaneously dissolved with the same 
universe and secondly, there is no end to the number of uni¬ 
verses. So, somewhere or other, there are always substrata 
for the universals.] 

Remark: The Mimamsakas do not believe in Mahapralaya. 

What is the answer of those who believe in the Great 

43. Those who believe in Unity hold that, once 
the different objects are dissolved in the ultimate cause 
(Prakrti), the universals have their existence as one 
with the ultimate substance. 

VAKYAPADIYAM of bhartrhari 

[Those who believe in the Great Dissolution must natu¬ 
rally answer this question differently. They say that what 
is called the Great Dissolutoin is nothing more than the 
merging of everything in the root-cause, the Great Universal 
(mahasamanya). The universal also merge in it which is 
here called substance (dravya). This root-cause is nothing 
more than the Great-Universal. According to Advaitins it 
is nothing more than the cause which is found in all effects 
. Vai;se5lkas believe that, in the Great Dissolution the 
universal exist without any substratum. But here* the 
author has stated his own view]. 

‘The author now makes a statement on the basis of the 
view that all universal exist everywhere. 

44. Universal like brahmana-hood reside in all 
living beings. Once they are manifested, they produce 
their effects. This is another tradition. 

[Universal like brahmana-hood exist everywhere but 
are manifested only through particular substrata. So’ one 
cannot argue that if the universal exist everywhere, jar-ness 
would exist in a cloth also and thus confusion would result 
Only a jar can manifest jar-ness and not anything else. Thus 
according to this view, all universal exist in the ultimate 
cause like the atoms at the time of the Great Dissolution j 

45. Some declare that the universal are mani¬ 
fested in pictures etc. But they become the cause of 
the attainment of merit and demerit only when they 
have living beings as substrata. 

[Some go to the extent of saying that the universal of 
sentient things exist in their non-sentient images. But this 
is not correct. All that happens is that, due to similarity 
between the original and the image, similarity of cognition 
takes place. That does not prove the presence of the 

m. 1.46 33 

The author now states how the omniscients acquire their 
knowledge of the universals. 

46. The knowledge regarding the universals of 
those who are different from us proceeds from all the 
senses, just as that relating to the characteristics of 
precious stones and coins comes through practice in the 
case of those who know them. 

[All words denote universals which exist in the parti¬ 
culars that manifest' them. But here a doubt arises. If the 
particulars manifest the universals, how is it that they do 
not do so for one who does not know the language. Such a 
person, after seeing a large number of cows and other ani¬ 
mals, does not seem to see the ‘cowness’ which exists in all 
cows and does not exist in the other animals and give expres¬ 
sion to it by the word ‘cow’. If the individual cows do mani¬ 
fest ‘cow-ness’, how is it that he does not perceive it? And 
if he ultimately comes to see it through others, how did the 
latter know it? Such reasoning will ultimately take us back 
to one who is naturally omniscient and knows all the uni¬ 
versals and their words and teaches others. The assumption 
cf the existence of such a person or persons becomes neces¬ 
sary for the knowledge of such universals as ‘brahmana-hood’ 
which differs from such universals as ‘cow-ness’ in that the 
latter can be observed by all as existing in all cows and 
not existing in other animals, whereas the former cannot' be 
so observed. These omniscient beings were the original 
teachers of mankind and they were taught by God ( Isvara ). 
As the Yogasutras say: Sa purvesdm api guruh, K&le- 
ndnavaccheddt. (Y. Dar. I. 26). When, after the Great Dis¬ 
solution, Brahma and others are created again, it is isvara 
who teaches them and they pass on the knowledge to others. 
All the sdstras testify to the existence of omniscient Beings, 
Isvara and the deities. They directly see the universals and 
the substrata which manifest them. They differ from us in 
that their senses are not restricted in their scope. They can 
hear with the nose, see what is behind them and do the work 
V-III. 5 

24 vAkyapadiyAm of bhahtrhaki 

of all the senses with the tips of their fingers. For them, 
knowledge of Brahmana-hood is not merely visual, but pro¬ 
ceeds from their other senses also. They can see subtle uni- 
versals in words. Besides Tradition, there is another proof 
for the existence of omniscient beings. We do see in life 
that practice increases both knowledge and power in some 
individuals. Jewellers acquire great skill in detecting the 
genuineness of precious stones. Whatever has degree reaches 
its climax somewhere. That Being in whom knowledge and 
power reach their climax is Isvara. From Him, some Yogis 
get their knowledge. These Yogis are superior to us. They 
can see what we cannot. They can see Brahmana-hood 
directly, just as we can see ‘cow-ness’. It is they who teach 
us that such and such a word denotes such and such a uni¬ 
versal. In doing so, they only make known what is already 
there. The relation between word and meaning is eternal.] 

The author now considers whether, in words like jati- 
gandhatva, ‘utpalagandhatva’ etc., the abstract suffix-rua ex¬ 
presses the wider universal, namely, the fact of being smell 
or the lower universal, namely, the fact of being the smell 
of j ati flower etc. 

47. The particular fact (smell) which is resorted 
to in words like jatigandha and utpalaganclha , is not 
conveyed in the world by the abstract suffixes, it being 
(relatively more) transitory. 

[It was said before that all words denote universals. Here 
a question arises: in words like j atigandhatva, utpala¬ 
gandhatva etc., dees the suffix-tua denote the universal of 
smell or the universal of a particular smell? The answer is 
that the suffix expresses smell in general and not the fact of 
being a particular smell, the smell of a particular flower. 
That is because the former is wider and more comprehensive. 
The quality of being a particular smell is relatively ‘anitya 
less comprehensive. So the suffix must be taken to ex¬ 
press that which is relatively more permanent and more 

III. 1.48-49 


The author now point's out that some universals have no 
words to express them. 

48. Universals like the fact of being a ‘man-lion’ 
have no words of their own to express them. They are 
conveyed by other words the (fictitious) parts of which 
are similar (to other words in the language). 

[Universals like ‘the fact of being a man-lion (nara- 
simhatva ) have no words to express them. The word nara- 
simha has two parts, each of which is like another word in 
the language. The result is that each of these words brings 
to our mind its own universal. The word nam brings the 
fact' of ‘being human’ to the mind and simha brings to the 
mind the fact of ‘being a lion’. But the universal ‘the fact 
of being a man-lion’ is something quite different from naratva 
and simhatva. It is not a mixture of the two. The word 
narasirnha is really an indivisible word and to say that it has 
two parts which resemble two other words in the language 
is really a fiction. Because the two fictitious parts of the 
word bring to our mind two other universals, the word can 
cause only an error in us. The fact is that such universals 
have no words of their own to express them.] 

The author now states that like the stem, the suffix also 
denotes the universal and that the two are coordinated. 

49. In any consideration of the scope of the 
universal (of a thing or of an action) the number or 
the universal in it (denoted by the suffixes) is of use 
because of its connection, even though it may be 

[The universal exists in a thing or in an action by the 
relation of inherence. The determination of its scope is based 
on the consideration whether it' resides in one thing or in 
more than one thing. When such a consideration is made, 
the universal of the number conveyed by the suffix in the 



noun or the verb plays a useful part. It does so, not because 
it is directly connected, but by the relation of ekdrthasaraaveta- 
samavdya = inhering in what inheres in a thing. Between 
the universal of number conveyed by the suffix in a verb 
and the universal of action, there is this relation through 
their substratum. The universal conveyed by the stem and 
the particular number reside in the same thing. Only that 
which has been understood can qualify, it does not matter 
whether it is understood directly or indirectly. According 
to the view that all words denote the particular, the parti¬ 
cular number is conveyed directly by the word and not in¬ 
directly understood. In such expressions as gauh, ga-vau 
pacati, pacatah, a thing having a particular number is under¬ 
stood as the main thing. Thus the meaning conveyed by the 
suffix becomes the visesana and the meaning conveyed by 
the stem becomes the visesya. This is also in keeping with 
the well-known saying that a verb is a word which primarily 
expresses an action. In the word pacati, for instance, the 
action of cooking which is the meaning of the root' is the 
main thing and it is qualified by the accessories etc. which 
are conveyed by the suffix. In the word aupagavci, on the 
other hand, the idea of progeny, conveyed by the suffix, is 
the main thing and it is qualified by the meaning of the stem. 
It is in reference to words like that that the saying Prakrti- 
pratyayau pratyaydrtham saha brutali' applies.] 

The author now points out that, sometimes, the meaning 
of the suffix is not meant to be coordinated with that of the 

50. Even when the powers of number, the accu¬ 
sative case and the like have the same verbal element 
to express them, it is seen that there is secondary usage 
( laksana ) when they (are not actually meant but 
merely exist to) give a certain completeness to the word 
or that they help in the accomplishment of the action 
(which is the real meaning of the sentence). 

III. 1.50 


[Thus, speaking generally, one can say that the word 
conveys the meaning of the stem as qualified by that of the 
suffix. That is how the meanings of the two elements of a 
word, both universals, are co-ordinated. Sometimes, things 
happen differently. That is, though the suffix is used in 
order to lend a certain completeness and correctness to the 
word, its meaning does not play any part'. In the sentence 
graham sammdrsti, the second case-affix expresses the singu¬ 
lar number, but it is not to be taken seriously, because more 
than one vessel ( graha ) is wiped. Similarly, in the sentence 
saktun juhoti, the accusative case expressed by the second 
case-affix in sciktiin is not’ to be taken seriously. All that t e 
sentence means is: the sacrifice is to be performed with ground 
grain (saktu) . The act of performing a sacrifice is meant to 
serve an invisible purpose and, therefore, that is the impor¬ 
tant thing. The homa is not performed in order to reduce 
the flour to ashes. It is performed in order to attain an 
invisible fruit, with, of course, ground grain as the material. 
Somebody at this stage might argue as follows. If this is the 
case, if is wrong to say that the accusative case in ‘saktun’ 
is not significant. One can speak of something which actual¬ 
ly exists as not seriously meant. For example, in graham 
sammdrsti, the second case-affix does convey the singular 
number which, therefore, really exists, but is not significant. 
But nobody says that' the accusative case is not' significant in 
K&sthdni pacanti=the fuel cooks, because it does not exist 
at' all. There, the nominative case is used instead of the 
usual instrumental case which, therefore, can be said to be 
not meant. But this line of argument is not right. Because, 
if, in the sentence saktun pacati, saktu is not the gramma¬ 
tical object (karma), that is, if it is not what the agent wishes 
to reach most, the invisible purpose itself cannot be at'taine . 
What one means by saying that the accusative case is no 
significant is that the sacrifice is not performed for the sake 
of ground grain. It is the latter which is made use o m 
order to attain some other purpose. 

Sometimes number and the accusative case sei\e the 
purpose of accomplishing the action which is the meaning 


vakyapadiyam of bhartrhari 

of the sentence and that is done by specifying in some way 
the things which are the accessories of the action. For ex¬ 
ample, in the sentence, pasund yajeta =one should perform 
the sacrifice by means of an animal, the singular number in 
pasund is significant', because the sacrifice can be fully ac¬ 
complished even with one animal. Similarly, in Vrlhhi ava- 
hanti =he threshes the paddy grains, the accusative case ex¬ 
pressed by the second case-affix is significant, because vrllii 
is what the agent wishes most to reach, because it' is to be 
cleaned by means of threshing. Threshing is never for its 
own sake, but' for the sake of the paddy-grains. When 
threshed, they become fit' to make the sacrificial cake ( puro- 
ddsa). Thus, they become accessories to the sacrifice. The 
grains are more important than the act of threshing. They, 
when threshed, are of further use Saktu, on the other hand' 
after it is thrown into the fire, is not of further use, because 
it exists no more. When it' is said that threshing i s less im- 
porant than the grains, it is from the point of view of rea¬ 
lity. As far as the language is concerned, it is threshing 
which is enjoined in the sentence vrllun avahanti .] 

The _ author now points out the scope of secondary power 
(lalcsand) in the formation of words. 

51. No finished object ( sattvabhuto’rthah) can be 
expressed without a number. Therefore, the number 
which is found in all expressions is not (necessarily)’ 

[A thing is what can be referred to by a pronoun such 
as ‘this’ or ‘that’. It is always associated with some num¬ 
ber. It cannot be otherwise expressed in language. So the 
number which a word expressive of a thing (dravya) has 
should not' always be taken seriously. It is there only to 
lend a certain completeness or correctness to the word]. 

To this general rule, some point out' an exception. 

ill. 1.52-54 


52. According to some, the singular and the plu¬ 
ral numbers are not significant, as they serve only to 
convey the universal. But the dual number is signi¬ 

[According to some, the singular and plural numbers 
may not be meant to be taken seriously. They may be only 
the means to convey the universal, as in the sentences: 
brahmano na hantavyah=a brahmana is not' to be killed, 
surd na peyd= wine is not to be drunk, vrsalair na pravesta- 
vy am=sudras are not to enter. Where the singular num¬ 
ber is significant, the word may primarily convey che indi¬ 
vidual ( dravya ). The dual number, on the other hand, can 
never be used for conveying the universal primarily. It al¬ 
ways denotes the individual primarily]. 

The author now says that, sometimes, even the dual 
number is not significant. 

53. In such a sentence as: “if these two fall ill, 
this medicine should be given”, the number two belong¬ 
ing to the objects is not significant. 

[But even the dual number is sometimes not meant to 
be taken seriously, as, in the sentence: ‘if these two (dual 
number) fall ill, this medicine should be given to them’. 
Even if only one of them falls ill, the medicine is given]. 

The author now points out that the number expressed 
by the stem is always significant. 

54. In sentences like: “he digs to the accompani¬ 
ment of two (mantras)”, the number expressed by 
words like ^one' etc. is to be considered as part of tne 
action, because this (number) arises out of the stem 
of the word. 



[The universal of number conveyed by the stem is al¬ 
ways seriously meant. In such expressions as dwbhy&m 
mantrabhyam. mrdam khanati, caturbhir ddcitte, astabhir 
harati, the number conveyed by the stem is significant. 
Otherwise, there would be no meaning in using that parti¬ 
cular stem at all. That particular stem denotes a particular 

It is now pointed out that the singular number, conveyed 
by the suffix, is also sometimes significant. 

55. „ In the sentence “one should sacrifice with an 
animal”, though there is a certain completeness (sams- 
kdra), the number one must be considered to be as 
much a part of the action to be performed as the 

[It was said before that the dual number is always seri¬ 
ously meant. It does not come in merely to lend a certain 
completeness or correctness to the word. A word which ends 
in a dual-suffix denotes primarily the individual and not the 
universal. But, sometimes, the singular number is also sig¬ 
nificant, as in the sentence: pasuna yajetci. Here the univer¬ 
sal conveyed by the stem, namely, that of an animal is an 
important element in the action. Similarly, the singular 
number conveyed by the suffix is also an important element. 
The two elements are conveyed by the same word, aooear as 
one in the mind and cannot be separated. The inner rela¬ 
tion between the different elements of the same word is 
apauruseya, not made by man. The injunction to perform 
the sacrifice, expressed by the sentence pasuna yajeta applies 
to the singular number also, through updddna or inclusion 
or implication which is different from the six authorities 
sruti etc. The injunction cannot, relate to a bare animal 
without any number]. 

Remark: See Mi. Su. 3.3.14. 

The author now says that, even those who hold that the 
suffix expressive of the singular number is only for the sake 

III. 1.5G-57 41 

of lending completeness to the word, consider it to be signi¬ 
ficant for other reasons. 

56. The mention of second etc is an indication that 
the singular number is significant. It is also an indi¬ 
cation that both the universal and number are found 
in the same thing. 

[After having declared: Agneyamajamagnistoma ala- 
bheta, it is said: aindragnam dvifiyam ukthye , aindrarr 
prsrtim tr&yayi soclasini. The very mention of dvitlya— 
second, trtiya=i hird etc is a proof that the number one 
expressed by the suffix is significant. It also proves that the 
universal and number are found in the same object. Unless 
the universal is found completely in one individual, it would 
be impossible to sacrifice a second animal. Unless the whole 
universal is found in an animal, it would not be that animal 
at all. As Helaraja puts it: Yavata hi vina nopapadyate 

57. (In the sentence ‘pasana yajeta) the animal is 
taught for the first time as the means of sacrifice, not 
taught elsewhere and also that it is subsidiary to the 
act of sacrifice. That is why it (the singular number) 
is significant. 

[The reason why the singular number is significant in 
the sentence pcisund yajeta is that it mentions for tne first 
time that the object to be sacrificed is an animal which is, 
therefore, an anga of the sacrifice. There is no other text 
where the number of animals to be sacrificed is mentioned. 
From this very text, we have to understand the number 

The author now says that the sentence graham sam- 
mdrsti is a different case. 

v-m. 6 



58. In such sentences as “vessels dedicated to 
Prajapati are nine in number”, the vessels are taught 
in different numbers. 

59. How can the number mentioned in connection 
with what is (generally) understood as being subsi¬ 
diary but becomes important as far as cleaning is con¬ 
cerned be significant? 

[This is not the case in the sentence ‘graham sammdrsti'. 
Here the singular number in ‘graham’ is not significant 
because there are other texts where the exact number of the 
vessels (grdha) in the different ceremonies is mentioned. 
In those texts, the number comes as the meaning of the 
stem. Here, the main purpose of the text is to teach the 
cleansing of the vessels. Thus, the vessels become, in rea¬ 
lity, the important thing ( ahgin ) in regard to the act of 
cleaning, though, verbally, cleaning is the important thing. 
It is natural that all the vessels previously mentioned should 
be cleaned and not merely any one of them. Therefore, the 
singular number is not significant for two reasons: (1) the 
real number is taught elsewhere (2) in regard to cleaning, 
the vessels become ahgin]. 

The author now says that the singular number becomes 
significant where there is a special reason for it. 

60. Neither the idea that the singular number 
lends completeness to the language when number has 
been taught elsewhere nor that it becomes important 
is the reason for considering it as insignificant. It is 
significant when an effort is made to make it so. 

[Though the above two reasons have been given for 
holding the singular in graham sammcirsti to be not signi¬ 
ficant the real reason is the one given in verse 51. It be¬ 
comes significant only if there is a special reason for it. There 
is no such reason in this case]. 

III. 1.61-64 


That' is what the author now says— 

61. No distinction has been made as far as the 
cleansing of vessels is concerned. They are taught (in 
different numbers) and they have to be purified. 
Therefore, all are taken. 

[There is no indication anywhere that only one parti¬ 
cular vessel is to be cleaned. The singular number in 
graham is no such indication as it' does not specify any 
particular vessel]. 

The author now points out what indication there is in 
'pasand yajetci. 

62. As the universal is present in full in every 
-substratum, one would think of resorting to more than 
one animal only if the act cannot be accomplished 

with one. 

[In the sentence pasuna yajeta there is a special reason 
for considering the singular number significant. If the uni¬ 
versal of the animal were to reside completely only in more 
than one animal taken together, it would be necessary to 
sacrifice all of them before the whole of the universal is sacri¬ 
ficed. That is not the case. The whole of the universal is 
found in any single animal.] 

63. When the act has been accomplished with one 
animal, if one takes another merely because it is possi- 
fcip to do so, such an act would be useless. 

64. Just as it is useless to sow the seed in a woman 
in whom it has already been sown, in the same way, 
it would be useless to resort to another animal when 
the act has been completed with one, 


[The sacrifice is complete when one animal has been sac¬ 
rificed. By sacrificing more than one animal, the sacrificer 
is not likely to get any extra fruit. Thus, it is simpler to 
say that the singular number is significant.] 

Others give some other reason than simplicity for con- 
sideling the singular number to be significant 

65. As the act can be accomplished by that much 
(that is, by one) some think that there is no reason for 
by-passing the number one. In this way, the singular 
would not be (openly taught as) accessory to action. 

[Some argue that we come to many only after havin«* 
passed one and as the sacrifice can be performed with one" 
we take one and not more than one. There . ’ 

reason for by-passing the number one. The number one 
would serve the purpose (sdvmrlhya ). Thus, due to simnli 
city and because it would serve the purpose, the singJL 
number is significant.] 

66. The indcations in ‘second’ etc. only confirm 
what has already been understood through reasoning 
From them, number is not understood as an accessory 
3S the universal is. 

[The words dvitlya, trtiya etc. are lingas, indications of 
what has already been understood through simplicity and ful¬ 
filment of purpose (laghava and sdmarthya). They say noth¬ 
ing new. The singular number in pasund does not say that 
it is also a part of the accessory, because it has already been 
understood through simplicity and fulfilment of purpose. 
Verbally, it is not significant and still it comes in as shown 
above and the work is done.] 

The view expressed in verses 65 and 66 was that of ‘some’ 
(kecit ). It is now criticized. 

III. 1.67-68 


67. If, on the basis of agreement and difference, 
number is actually understood from the text, it is right 
that it should also become an accessory and not merely 
be secondary to the meaning of something else. 

[This is not the right way of looking at it. The real 
position is that the singular number is understood only if 
the suffix of the singular is present and not otherwise. If it 
is thus understood, what is the reason for abandoning it? 
It should also be considered a sddliana, accessory to action. 
It is not there merely to give a certain completeness to the 
meaning of the stem. The word pcisund expresses two ideas, 
one through the stem and the other through the suffix, and 
they stand in the relation of primary and secondary to each 
other. To accept one and reject the other is not right. 
Secondly, unless the singular number is taken as actually 
taught, it would not be possible to substitute clarified butter 
when, for some reason or other, one of the avadana-s is lost. 
One can use another animal, because there would be no obli¬ 
gation to stick to one animal. But if the singular number 
is openly taught one cannot resort to another animal and so 
one can take clarified butter as a substitute. 

68. When the full meaning of the word (includ¬ 
ing both the unversal and the number) is the acces¬ 
sory to action, the power of the word to convey both 
is not affected and that is an indication that number 
has also a function to perform in regard to action. 

[When, both the universal and the number are equal 
because both are expressed by the same word, both become 
accessories. Thereby ^ the word does not lose its capacity to 
express the accessory. From this we understand the func¬ 
tion of number. As smoke is invariably concomittant with 
fire, it causes the inference of the latter. Similarly, because 
the power of the word to convey the accessory is not lost, 
so it becomes the cause of our understanding the function of 
number. Where there is no such indication of the function 



of number as for example, in graham sammarsti, there the 
number is not significant.] 

The author now adopts another argument and reaches 
the same conclusion. 

69. The unknown ( apurva ) is to be considered 
important as it is to be taught; the known ( vihita ) is 
to be considered secondary, as it exists for something 

70. As (in the sentence in question) cleansing is 
to be taught and considering that the vessels are taught 
elsewhere (for the first time) the number found in the 
injunctive sentence is not to be rejected in (the inter¬ 
pretation) of the confirmatory sentence. 

[The question which is being discussed here is when the 
meaning expressed by the suffix, such as number, is signifi¬ 
cant and when it is not. So far, the question was discussed 
on the basis of the nature of the meaning conveyed by the 
suffix. Now, it is being discussed from the point of view of 
the function of the sentence. That which is vidheya, that is 
to be primarily conveyed by the sentence is the new feature 
and that is the important thing. That which is already made 
known elsewhere is mentioned in order to convey some¬ 
thing else through it. It is only restated ( anudita ). In the 
sentence graham sammarsti, it is the cleaning ( sammdrga) 
which is the new thing to be taught in regard to the vessels 
already taught elsewhere. In that other passage, the num¬ 
ber of the vessels is also mentioned. That number is the 
one which is meant in the sentence in question. The num¬ 
ber expressed by the suffix here is not to be taken seriously. 
The suffix only serves the purpose of lending correctness 
or completeness to the form ( sabdasamskdra ). The number 
expressed by it is set aside by the number openly mentioned 
by the stem in that other sentence. Before, the line of argu- 

ill. 1.71-72 


ment was that the cleaning is for the sake of the vessels 
which are, therefore, more important in the sentence under 
consideration. Here, the line of argument is that what is men¬ 
tioned in the injunctive sentence is more important than what 
is mentioned in a mere re-statement ( cinuvdda ,).] 

The author now points out that the case is different in 
the sentence pasund yajeta. 

71. There is no contradictory number for the ani¬ 
mal in some other injunctive sentence; therefore, it 
(the animal) is understood together with its qualities 
and the singular number. 

[As to the sentence, pasund yajeta, it teaches sacrifice 
with the animal as its accessory (guna ). The animal is 
taught for the first time. Therefore, the particular number 
expressed by the suffix is also significant. Together with the 
animal which is specified ) it becomes an accessory of the 
sacrifice. This way of looking at it is in accordance with the 
view of the Mimamsakas. (See Mi. Su. III. 1.7.13-15).] 

The author now says something about other qualities 
besides number. 

72. As an action depends upon substance ( dravya), 
so does it on a quality which is taught in connection 
with an action whose relation to a substance is already 

[Just like number, other qualities also become accesso¬ 
ries to action through the process of specifying a substance. 
Action requires accessories and it is a substance which can 
be so. But a substance is the substratum of qualities. It 
has always some quality or other. So action depends upon 
qualities also. It is quality which specifies a substance which 
is the real accessory to an action. In regard to action, a 
substance is an inner (antaraiiga) factor, whereas quality is 



an outer (bahiranga) factor. That is why substance is con¬ 
nected with action first and, through it, quality also is con¬ 
nected. This sequence is significant. Substance is the sub¬ 
stratum of power and power is the real accessory to action.] 

73. Because of their mutual indispensibility a 
particular quality is associated with a substance. Simi¬ 
larly, a substratum is also invariably associated with a 

74. When, because of this mutual indispensibility, 
a relation between the two is understood, it should not 
be abandoned nor something not actually mentioned 
be brought in. 

[Substance and quality require each other. This mutual 
requirement is called samarthya. Granting this mutual de¬ 
pendence, when a relation between them is understood from 
the very words of the sentence, it cannot be abandoned. All 
this about substance and quality is being said from the point 
of view of artificial abstraction (apoddhara ). In reality the 
sentence is one and its meaning is an indivisible complex 
whole. Therefore, there cannot be any question of bringing 
in the meaning of a word which has not been mentioned in 
the sentence. As there is a mutual indispensibility between 
substance and quality, if both are expressed by separate ver¬ 
bal elements in the same sentence, they must be connected 
with each other. There is no justification to abandon this 
connection and connect any of the two with something not 
mentioned in the same sentence. A cognition which arises 
without any hitch must be taken as authoritative. The sen¬ 
tence is indivisible and conveys a single particularised mean¬ 
ing. For the sake of convenience in explanation it is divided 
artificially into parts and each part is assigned a meaning 
which is a universal. In association with the meanings of 
other words in the sentence, it becomes particularised. But 
this way of looking as it is only for the sake of convenience. 

111 . i.75-78 


Really speaking, the meaning of the individual word conti¬ 
nues to be a universal and only appears to be particularised 
in association with the meanings of other words. The word 
expressive of the universal disappears, as soon as it has been 
uttered, with its universal meaning and so with which parti¬ 
cular is it to be identified? The speaker cannot mean to 
express both the universal and the particular at the same 
time. Besides, the other words also express universals and, 
so, with which particular is each universal to be identified? 
Therefore the right view is that the individual sentence con¬ 
veys a particularised meaning. So there is no question of 
bringing in an idea not openly expressed in the sentence]. 

It is now pointed out that, even when two relations come 
out in the same sentence, there need be no ?;a/c yabheda, 
splitting of the sentence, because of the two relations. One 
can be the main thing and the other secondary. 

75. The relation of these two with the action is 
expressed by means of a verbal element (case-ending). 
The close relation between the substratum and that 
which exists in it is understood from the sentence. 

[In the sentence gam cibhyaja suklam dandena , both the 
substance and the quality stand in the relation of accessory 
towards the action and this is made known by the case-end¬ 
ing. The words gam and suklam stand in the relation of 
qualified and quality to each other and this is understood 
through the sentence. Thus, the sentence conveys two re¬ 
lations and yet it is one. There is no splitting of the sen¬ 
tence. Of the two relations, one is the main one and the 
other secondary to it. Thus, in one expressive movement, 
the sentence conveys both, that is, it conveys a particular 
substance, having a particular quality as the accessory to 

1 The author now points out a difficulty. 

76. In this view, when the (prescribed) substance 
or quality is not available, it would be possible to have 

V-m. 7 


VakyApadiyAm of bhartrhari 

a substitute for either. The relation expressed directly 
by a verbal element is stronger than the one conveyed 
by the sentence. 

[According to this view, there would be a difficulty. 
When, either the substance or the quality prescribed is not 
available, there could be a substitue for either, because that 
which is conveyed by a directly expressive verbal element is 
stronger than what is conveyed by the sentence. Therefore 
the injunction ‘One should sacrifice a white goat’ could be 
carried out with a black goat or a white goat made of flour 
if the prescribed substance or quality is not available ] 

If the universal is the meaning of words, evervtbino 
would be alright. S 

77. When the view is that it is the universal or 
capacity which is taught in relation to action, then 
substance and quality are also understood because of 
their indispensibility. 

[Accordingly to the view that the universal is the mean¬ 
ing of words, everything would be alright, because, then the 
connection of both substance and quality with action would 
not be direct, that is, through the verbal element. Only th~ 
connection of the universal or of the capacity with action 
would be so. It is the universal which is taught as the acces¬ 
sory to action and it has been declared that the universal 
stands for that which has the requisite capacity. So it is 
the connection of the universal or that of capacity which is 
directly expressed by the verbal element ( sruti ). But both 
must have a substratum and so a substance is also understood 
and through it, a quality also.] 

78. Even though universal and qualities are 
equally accessories of action, it is the qualities of the 
goat etc. for which there may be substitutes and not 
for the universals. 

III. 1.79-80 


[Both the universal and qualities enter into relation with 
action through a substance and yet, there cannot he a sub¬ 
stitute for the universal. Only a quality can be replaced.] 

The author now says why. 

79, The qualities are not as near to the substance 
having capacity as the universals are. The substance 
is directly connected with action. That is why there 
can be alternative only for quality. 

[That is because the universals are much nearer to the 
substance which is actually the accessory than the qualities 
which come later and are liable to change. So they are a 
step removed from substance which is directly connected with 
action. No quality can be connected directly with action. 
As the universal is much nearer to the substance which is 
connected with action, there can be no substitute for it. If it 
is replaced, it would become a different accessory and so 
the action itself would become different. All this is said on 
the basis of the universal being the meaning of words. Accord¬ 
ing to the view.that substance or the individual is the mean¬ 
ing of words, the universal is not directly taught and so can 
be replaced.] 

80. Some (on the other hand) think that, because 
of resemblance, there can be alternative for either: 
there can be a goat which is not of the (prescribed) 
colour or a sheep which is. 

[The oilier view that there can be a substitute for both 
the universal and quality, because both are connected with 
action through substance, is not the view of the Bhasyakara. 
but of others, of those who follow Vyadi.] 

After this digression, the author comes back to the sub¬ 



81. It is seen that the universal, together with the 
number attached to it, is applied (to action); some¬ 
times, the universal alone, after discarding the num¬ 
ber, is applied. 

82. Where the universal is subsidiary to something 
else, there it is associated with the number of the 
substance (individual). Where, on the other hand, it 
exists on its own, it does not depend upon the number 
(of the substance) . 

[Before it was declared that the singular number in 
graham is not significant on the basis of verbal presentation 
(sabdavydpara) that is, whether something is presented by 
words as the main thing or as the secondary thing. Here it 
is said on the basis of the factual situation ( vastavam ), that is. 
whether something is actually the main thing or the secondary 
thing, that the number of the latter is significant.] 

83. In the sentence ‘one should sacrifice with an 
animal’ (yajeta pasuna ), the mention of animal is for 
the sake of the sacrifice and the act, (the sacrifice) the 
main thing, is fulfilled even with one animal. 


[In the sentence pasuna yajeta, the sacrifice is the impor¬ 
tant thing. The animal is an accessory to it. As the sacrifice 
can be accomplished with one animal, there is no reason for 
sacrificing a second one. The universal of the animal, being 
subsidiary to the sacrifice, is associated with the singular 
number expressed by the suffix in pasuna. If a second ani¬ 
mal is sacrificed, there would be another act altogether, which 
is not pi’escribed. Number should follow the main prescrip¬ 
tion and not vice-versa. If one were to sacrifice as many 
animals as one can, the sacrifice would follow number and 
not vice-versa.] 

III. 1.84-8G 


81. If one were to sacrifice with as many animals 
as one can, then it would be a case of the main thing 
being determined by the subsidiary things. 

The author now gives an example of the principle refer¬ 
red to in stanza 82. 

85. Where the act (of cleaning) is mentioned, the 
main thing being the vessel which is cleaned, there we 
disregard the particular number which is expressed. 

[Where the vessel is to be cleaned, it becomes the main 
thing in relation to the act itself. There, the number ex¬ 
pressed by its own vevbal element, the suffix, is not signifi¬ 
cant. If the act were the main thing the number one expres¬ 
sed by the suffix would be significant, because the act can 
be accomplished even by cleaning one vessel. Cleaning more 
than one would be useless. But that is not the case.] 

The author now points out an indication of this princi¬ 
ple in the science of Grammar. 

86. The very fact that in the sutra teaching the 
retention (of a word as the main thing) the word one 
is specially mentioned, even though the word sesa (left 
over, retained) has a suffix expressive of a particular 
number (one) is a clear indication (of the principle). 

[The principle that the number expressed by the suffix 
in regard to the subsidiary is significant, but not the one ex¬ 
pressed in regard to the main thing can be known from the 
science of Grammar. In the sutra: sarupdndvi ekasesa eka- 
vibhakatau”=Of many words having the same form, only one 
remains when they are followed by the same case-ending 
(P. 1.2.64), there is the word ekasesah. The suffix in sesnli 
expresses the singular number. So does the word eka (one). 
The very fact that the word eka has been put in even though 
the suffix is there shows that the number expressed by the 



suffix is not significant. What remains is taught and is, there¬ 
fore, the main thing. Hence the number mentioned in con¬ 
nection with it is not significant.] 

An illustration of the principle from the science of Gram¬ 
mar is now being given. 

87. In the rules teaching compound words and 
suffixes, just as there is express mention (of a parti¬ 
cular number), in the same way, it is based on reason 
also, namely, that the secondary things (parts') depend 
upon something else (the whole). 

[In the sutra saha supd (P. 2.1.4.) the main thing is the 
compound word ( samasa ) which it teaches. It is not the 
members of a compound word which are taught in that sutra. 
Really speaking, a compound word like a sentence is indi¬ 
visible. So it has no real parts. Parts are derived by abs¬ 
traction. The elision of endings, shortening of final vowels, 
masculine form etc. which are taught in regard to the mem¬ 
bers of a compound word are really for the sake of the cor¬ 
rectness of the whole. In the sutra in question, the singular 
number is actually heard and it is justified by reasoning 
also. That is why, in a compound word, a word combines 
with one other word at a time and not with more than one. 
Similarly, in teaching suffixes also, the particular number 
mentioned is sometimes significant, as in the sutra nydpprdti- 
padikat (P. 4.1.1.). According to this, the suffixes taught in 
the subsequent sutras come after a single word ending in 
fit or dp and after a single stem ( pratipadika ). One suffix 
cannot come after many of them at the same time.] 

It is now shown that there are exceptions to this principle. 

88. (Sometimes) even in what is secondary, the 
number (mentioned) is not accepted (that is, it is not 
significant) in order that the main thing may be 
accomplished. Therefore, in regard to the teaching of 

111. 1.89 55 

the name ‘object’ (karma), the agent is understood un¬ 
qualified by anything. 

[It is not that, in the sastra, the number coming after 
every secondary thing ( gwci ) is significant. The sutra 
teaching the name karma is: kartur ipsitatamam karma— that 
accessoi’y is called the object which the agent wants most to 
reach (P. 1.4.49). Here the word kartuh is in the singular num¬ 
ber and in this sutra, the agent ( karta ) is secondary, because 
the name ‘object’ is the main thing. And yet, the singular 
number here is not significant. What many agents together 
want to reach most is also called karma.'] 

This conclusion is now further strengthened. 

89. When something which is possible is expressly 
mentioned in order to remove other possibilities, there 
the number is not significant, as its purpose has been 
served by the removal of other possibilities. 

[If Paiiini had only said Ipsitatamam karma=‘that which 
is most desired to be reached is karma,’ and not specified by 
whom, any accessory ( kdraka ) could occur to one. But as 
he has added kartuh (by the agent), the other accessories 
are excluded. This exclusion is the purpose of the mention 
of the agent in the sutra. The singular number of the suffix 
in kartuh is not significant. It has been used because, after 
all, some number has to be used. Even what is desired to 
be reached by more than one agent gets the name of 
karma. This reasoning cannot be extended to the case of 
pasuna yajeta, because that is the main injunction in regard 
to the animal and so the singular number in pasuna is also 

• i 

The author now explains how a single case-ending comes 
after many identical forms if the singular number in P. 4.1.1. 
is significant. 


VAKYAPADIYAM of bhartrhAr! 

SO, Where a case-ending comes after several 
identical stems, one of the latter is meaningful and is 
expressive of the whole group. 

[The retention of one only of several identical forms 
(ekasesa) is taught in P. 1.2.64 when the speaker wants to 
use them together (sahavivaksayam ). When the meanings 
of the different identical forms are mutually connected, the 
meaning of the group comes into being. The different iden¬ 
tical forms of the group and the group as a whole are mean¬ 
ingful and, therefore, get the name of stem (pratipadika ). 
There can, therefore, be a case-ending after the group also, 
even if the singular number in P. 4.1.1. is significant. Once 
there is a case-ending after the group, it will serve each indi¬ 
vidual form also. No separate case-ending need come after 
each identical form. Thus several identical forms would be 
followed by one case-ending, only one of them would remain 
and the others would disappear according to P. 1.2.64.] 

In this way, the main thing would be fully supported. 

9r. According to the view that the number ex¬ 
pressed by the word denoting what is secondary is 
significant, the main thing, whether it be a suffix (as 
in the sutra nyappratipadikat —P. 4.1.1.) or a com¬ 
pound word (as in the sutra sdha supa —P. 2.1.4.) is 
supported in every way. 

[Even according to the view that the number expressed 
by the word denoting what is secondary is significant, a suffix 
can be added after a single stem or a compound stem accord¬ 
ing to P. 4.1.1. and a compound word can be made by uniting 
one word with another or with many ) as in the case of a 
dvandva or a bahuvrlhi.'] 

The author now begins to state a few more points about 
the universal. 

iii. i.92-94 


92. Similarity consisting of absence of difference 
or the powers which are the very essence of things, 
these might be described as the synonyms of the 

[The universal is something the existence of which is 
inferred because of the similarity of the cognitions which the 
individuals produce. It is never seen apart from the indivi¬ 
duals. Why not accept that the individuals are themselves 
the cause of this similarity of cognition on account of their 
resemblance? What is the use of postulating a universal? 
The mutual resemblance of the individuals, not different from 
the individuals themselves, is the cause of the similarity of 
cognition. It is different from the expressed conventional 
meaning of a word, it is a kind of power. This is just what 
is called universal. After all, the universal is nothing more 
than something in the individuals which causes a similarity 
of cognition. On this everybody is agreed. That something 
can be called resemblance, power etc. The only point on 
which the different thinkers differ is: Is it something over 
and above the individuals?] 

The author now states that it is something over and above 
the individuals. 

93-94. From a man’s desire to hold a stick, one 
no doubt gets the idea of a stick, but still one does not 
get the idea that he actually holds it. From a man’s 
desire, one does not get the knowledge which has the 
form ‘he is one with a desire’. Therefore, even though 
there is power, the mind has recourse to something else. 

[On this point, the followers of the Vaisesika say that 
it is something over and above the individuals. They reason 
as follows— The universal is postulated because, the indivi¬ 
duals, being different from one another, cannot account for 
the similarity of cognition. Resemblance and power are not 
over and above the individuals. Therefore on® has to postu- 
V-III. 8 


vakyapadiyam of bhartrhari 

late the universal which is over and above them. These 
universal have particular abodes, - namely, the individuals 
having a particular disposition of parts. . A cow, having a 
particular disposition of parts is the substratum of cow-ness. 
The samething can be said of horse-ness. If a particular 
arrangement of parts determines the substratum of each 
universal, why not attribute similarity of cognition also to 
this very arrangement? Why postulate the universal, apart 
from the disposition of parts? The reason is that particular 
effects have particular causes. When we get about some¬ 
body the cognition that he is the holder of a stick ( dandl ), 
the cause of it is his connection with the stick. The cause 
of this connection is the desire of the person in question to 
take a stick. We cannot say that this desire is the cause 
of our cognition that he is the holder of a stick. Similarly, 
we have to postulate a special cause for the similarity of 
cognition which we have when we look at different indivi¬ 
duals belonging to the same class. Thus resemblance and 
the particular disposition of their parts can account for cer¬ 
tain things, but not for this. This can be explained only by 
the universal, which is over and above the individuals.] 


95. In the end, the indeterminate nature (of things) 
or power would remain and all verbal usage would 
become difficult. 

[If we do not accept direct and immediate causes for 
effects, but try to explain them by remote causes, it would 
take us back to the root cause of everything, whether it be 
primordial matter ( pradhana ) or anything else. It would 
be something indefinable. It would, then, become impossi¬ 
ble to make statements of cause and effect. To explain 
verbal usage, universals have to be accepted.] 

* * o 

The author now states that, according to the Sankhya 
and the Advaita, the universal is not something over and 
above the individuals. 

III. 1.96-98 


96. When, abandoning all distinctions, the essence 
of the individuals is perceived as one, then a single 
conception comes into existence. 

[We trace back the whole universe to one cause if we 
find some common feature in everything. That which does 
not exist in any way cannot come to be. Thus cause and 
effect are really the same thing. Everything is like every¬ 
thing else and yet not so. As one and the same thing 
becomes many, the common point is the universal and the 
difference is the distinguishing feature. The most compre¬ 
hensive common feature would be the great universal 
(mahasamanya ) and the less comprehensive ones would be 
the intermediary universals (avantarasamanya) . The factors 
that determine the parts are also transformations of the one. 
Everything is a transformation ( parinama or vivarta). Thus 
there is no absolute difference between cause and effect, 
dharma and dharmin, sdmdnya and uisesa.] 

As the siitra teaching retention of one only of many iden¬ 
tical forms has been declared to be impossible if the universal 
is the meaning of words, the author now brings it within the 
sphere of the idea of collection. 

97. When, that which has diversity is looked upon 
as one then from the many is born the conception of 
a ‘collection’. 

[When both diversity and unity figure in the cognition, 
then the object is what is called a collection. Unity is super¬ 
imposed on what is diversity. The cognition of collection 
arises from the many, to which unity is subordinate. In the 
cognition of the universal, diversity does not figure. Unity 
or identity is the predominant factor.] 

The author now defines the scope of resemblance. 

98. When the difference between the different 
individuals is kept down and the two (unity and diver- 



sitv) are understood together (sakrtpravrttau) there 
is the idea of collection, and when they are understood 
in turn, there is the idea of resemblance. 

[Resemblance, collection and the universal are allied 
things, but there is difference between them. In the 
cognition of resemblance, both difference and identity figure. 
In the cognition of the universal, only identity figures. In the 
cognition of collection, unity mixed up with diversity figures. 
The two appear at the same time. In the cognition of resem¬ 
blance, on the other hand, first difference figures and then 

The author now explains the Bauddha view in regard to 
the universal. 

99. Just as a cognition which is dissimilar appears 
to be uniform, similarly, an object also appears to he 
the same even though it is different. 

[According to the Bauddhas, the cognition of Univer¬ 
sal is an error. All cognitions are different from one another. 
There is no such thing as identity of cognition, but, due to an 
eternal vdsana, cognitions appear to be identical and on that 
basis, the concept of universal in cognition arises. Objects are 
also essentially different from one another, but due to the 
same vascina, they appear to be identical and this unreal iden¬ 
tity is called the universal.] 

The author now considers whether there is any external 
basis for this cognition of identity. 

100. Nor are there words exoressive of the 
difference between objects which look like one another 
nor is the difference itself perceived. 

101. Because of the difficulty in perceiving the 
difference in our cognitions, words and objects, we 
perceive them to be identical. 

III. 1.102-103 


[Things are essentially different from one another. 
There is nothing in common between them. But we are not 
able to see this difference. As our words can only express 
what we see, they also do not express this difference. They 
express a certain identity which we see on account of re¬ 
semblance. Thus, our cognitions of things (prakhya) and 
our expression of them by means of words ( updkhyd ) pio- 
ceed on the basis of an unreal identity based on resemblance. 
The meanings of words are based on convention and con¬ 
vention cannot rest on the own nature of things ( svalaksana ). 
Cognitions really differ from one another, because their 
Qj^jgcts are different. Words also differ from one anothei, 
even when they are perceived as identical. Objects of the 
same class also differ from one another, but we notice only 
the point in which they differ from things of another class. 
The subtle differences which exist in precious stones, pearls, 
corals etc. can be seen only by experts and not by every¬ 

Those who accept universals, however, declare— 

102. There are universals in all cognitions as in 
objects and they prove the existence of universals i’i 
objects—such is the view of those who believe in con¬ 
nection. ( samsarga ). 

[Those who believe in universals believe that there are 
universals of cognitions also. It is the existence of universals 
in our cognitions which prove the existence of universals in 
the objects of these cognitions. This is one view of those who 
maintain that objects are connected with difference and 
identity ( samsargadarsana) ]. 

The following is another view current in the same 

103. The universal which exists in the object 
serves the cognitions also. Cognition is not determin¬ 
ed, like the object, by something which is external. 



wll .[ USt , a ff S the ° bject is 2 iven a form by the universal 
which is different from it, in the same way, one cannot say 
that cognition also is given a form by the universal which 
different from it. Cognition never becomes an object If 
one seems to perceive a universal in the cognitions, it is only 
due to the universal in the objects. Therefore, cognitions are 
endowed with a form ( sdkdrdh ), self-illuminating (svapra- 
Kcisali) and devoid of universal (nihsamanyah ).] 

How is cognition not determined by something external? 

104. Just as light is not illuminated by another 
light, in the same way, the form of cognition is not 
determined by another cognition. 

[A light which illuminates an object does not require 
another light to illuminate it. It is self-luminous. If the 
light is not self-luminous, it would not be able to illuminate 
the object either. The Vaisesikas also accept that the uni¬ 
versal in the object serves cognition also, but they do not 
believe in the self-luminousness of cognition. But the gram¬ 
marians do not follow them in this respect.] 

The author now points out the difference between the 
knowledge of an object and the knowledge of a knowledge. 

105. The knowledge “this is a knowledge of the 
jar is different from the knowledge “this is a jar”. 
The knowledge “this is a jar” refers to an external 

[Here a doubt arises: We can have two cognitions taking 
the forms ‘this is a jar’ and ‘this is a knowledge of the jar.’ 
In the former, the jar is the thing known. In the latter, how¬ 
ever, the knowledge of the jar is not the thing known, be¬ 
cause no knowledge can become the thing known. Know¬ 
ledge illuminates something else and is self-luminous, but 
never the thing illuminated. This argument is not valid, 
because knowledge is not self-luminous. It has to be illumi- 

nated by another knowledge. Nor need this lead to regres- 
sus ad infinitum 

The above doubt is answered as follows— The knowledge 
This is a knowledge of the jar’ is different from the know¬ 
ledge “this is a jar”, because the former is not produced 
directly by an external object. The jar which seems to 
figure in it has no reference to an external object. In other 
words, no object, different from external object, figures in it. 
Knowledge itself does not figure as an object.] 

106. As the form of the knowledge is not cognised 
as an object, its form as distinct from that of the object 
is not grasped. 

[Wny not accept that in the knowledge This is a know¬ 
ledge of the jar’, just as the form of the external object figures, 
in the same way, the form of the knowledge itself, mixed up 
with the former, also figures? If that is so, it would mean 
that knowledge also becomes the object of knowledge. But 
this position is not valid. The external object figures in this 
knowledge, but not the knowledge itself, as the illuminator 
of the external object and as distinct from it. Therefore, 
what is regarded as the knowledge of a knowledge is really 
the knowledge of the external object that figures in that 
knowledge. A cognition having the form of another cogni¬ 
tion figuring in it as its object does not arise. There can be 
no cognition which does not have an external object figuring 
in it. So ? in the cognition gliatam aliavi janami, it is the 
jar which figures on the object, not the knowledge of it. 

Thus, it has been shown that the word conveys the 
universal which is capable of accomplishing worldly usage 
relating to visible and invisible things.] 



Section 2 


Following Vyadi, the author now states that substance, 
which is the thing qualified, is the meaning of words, by first 
of all giving its synonyms. 

1. The Self, the thing-in-itself, Being, the Body 
(Primordial matter), the Elements, these are synonyms 
of the word Substance and it has been declared to be 

[According to Vajapyayana, the universal is the meaning 
of every word and it figures in the mind as the attribute 
(visesana) of substance which is also understood from the 
word by implication. According to Vyadi } substance is the 
meaning of every word and it figures in the mind as the thing 
qualified ( visesya ). It is this which plays the chief part in 
purposive action and it is that, therefore, which prompts one 
to act. The word primarily conveys that. The universal is 
not conveyed by the expressive power ( abhidha ) of words, 
but it specifies or determines the substance conveyed by it. 
Substance is of two kinds: real and expressional (paramar- 
thika and samvyavaharika) . It is the second which, accord¬ 
ing to Vyadi, is the meaning of all words or rather all things 
can be presented by words as substance. Here, in this chap¬ 
ter, we are concerned with the first kind of substance. It is 
called by different names in different systems. The monists 
call it atma, the Self. The same Self appears as different 
things through different limiting factors (upadlii) which are 
the immediate meanings of the different words. According 
to the Bauddhas, the thing-in-itself ( svalaksana ) is the real 
substance. Followers of sattadvaita consider that substance 
is nothing more than the Being which is the own essence of 
a thing ( svo bhdvah ). When inner sequence is not meant to 
be conveyed Being ( satta ) becomes a thing ( sattvci ) and 
this, differentiated by different limiting factors, in substance. 

III. 2.1-2 


For others, the body or primordial matter ( prakrti ) is sub¬ 
stance. For the Carvakas, the four elements, air, fire, water 
and earth, are the substance. They call it the reality (iatlva). 
When these elements combine, the body, the sense and the 
object result. These words are synonyms of the word 
dravya, because they denote the ultimate substance, which 
words like jar ( ghata ) cannot do. They can be applied to 
anything as in the statement: eko’ yam dtmd udakain nama 
(M. Bha. I. 1.1.). Here the word atma is used for water 
kept in a particular vessel. Other words denote substance 
through akrti. These, on the other hand, denote substance 
directly. Patanjali has declared in one place thcit while 
shape (akrti) constantly changes, substance remains the 
same. Thus, it is eternal. What the Bhasyakara means by 
eternal is the fact of something not giving up its essence 
even while forms are changing and, in this sense, dravyci is 
eternal even according to the Carvakas.] 

After having stated what substance is according to other 
systems, the author now shows the comprehensive nature of 
it according to Grammar. 

2. Through the unreal forms, it is the ultimate 
Reality which is cognised. By the words which directly 
express the unreal limiting factors, it is really the ulti¬ 
mate which is expressed. 

[Words are, by their very nature, incapable of express¬ 
ing the ultimate reality directly. So they are seen actually 
to express it thorgh various limiting factors. These latter are 
impermanent. They come and go. It is they which fulfil the 
practical needs of people. Words, therefore, express the 
forms which are cognisable. They are applied to things as 
they are cognised and they are cognised through their forms. 
A formless thing cannot be conceived by the mind. Even 
though words directly express these forms, they also denote 
the ultimate through these forms.] 

v-m. 9 


The author now shows how, if the words go beyond the 
limiting factors, the latter can be so considered at ,all. 

3. It is like the house of Devadatta being recognis¬ 
ed (or differentiated from other houses) by means of 
an impermanent feature of it and yet the word ‘house’ 
denoting only the bare house. 

[Even though a crow sitting on Devadatta’s house may 
help one in recognising it or distinguishing it from others, 
yet when one speaks about Devadatta’s house, the crow is 
not included in it. It was just a temporary mark which 
served a particular purpose. When it flies away, one can 
still recognise the house with the help of some more perma¬ 
nent feature like a dais ( vedika ) or white lotus ( pundarika) 
(See M. Bha on P. 1.1.26.), observed when the crow was still 
there. In the same way } the past passive suffix ta is disting¬ 
uished from any other ta‘ by the ‘k’ which is attached to 
it at the beginning. But that disappears in actual usage. 
Then, it is distinguished from any other ‘ta,’ by something 
more permanent like the accessory to action and the tense 
which it expresses. In the same way, words, while first ex¬ 
pressing a mere impermanent limiting factor, can point to 
a permanent reality.] 

As the crow is quite different from the house, it is 
natural that the house should not include it. The author, 
therefore, gives a better illustration. 

4. (Or) it is like gold etc. which even though 
differentiated by different impermanent forms, remains 
in its pure form, the expressed meaning of words like 
rucaka and so on. 

[Gold is one, but is differentiated by different forms such 
as ‘rucaka’ (the name of an ornament). These forms, how¬ 
ever, come and go, but the gold persists. The different prac¬ 
tical purposes are not served by the perishable forms. There- 

III. 2.5 


fore, words like ruccika do not express these forms but go 
beyond them and designate the permanent reality, that is, 
gold. Similarly, words express the ultimate reality by going 
beyond the plurality which is not ultimate. The imperma¬ 
nent forms are not the real expressed meanings of words. 
As they are not real, they cannot fulfil practical purposes. 
That is why they are not the expressed meaning of words. 
There is no harm in taking the view that words primarily 
denote the ultimate reality as visesya and the impermanent 
forms as visesaya. All that is emphasised here is that the 
impermanent forms or the limiting factors are not the real 
meaning of words.] 

The author now explains how, if every word ultimately 
points to the ultimate reality, confusion does not result in 

5. Just as the capacity of the eye etc. is limited by 
the tube etc., so is the capacity of words to convey all 
meanings restricted by the particular forms which they 
bring to the mind. 

[There would be no confusion because each word points 
to the substance Brahman through a particular form which 
that word and that word alone can bring to the mind. The 
word jar (ghata ), for instance, points to Brahman through the 
form of a jar. Just as one who looks through a tube sees only 
as much of reality as is visible through the hole of the tube, 
similarly, through each word, one can see only a particular 
form, limited by nescience and this form points to reality. 
Covers and impediments only restrict the power of the 
senses. They do not, in any way, affect reality. Similarly, 
nescience only restricts or limits the jivci-s, does not affect 
ultimate reality. It is not merely such things as tubes which 
restrict our vision. Attention concentrated on one thing takes 
it away from others. The particular beauty of a thing 
attracts our attention towards it and takes it away from 



The author now says something about words like sanni- 
vesa (=form, shape) which seem to denote attribute ( dhar - 
ma) only and not substance. 

6. As for the word which conveys such (imperma¬ 
nent) forms, since these are essentially one with it (the 
Substance), it also conveys the eternal, 

[Another name for the limiting factors ( upddhi ) is 
clharma (attribute). Words like sannivesa mean form ? shape, 
diposition of parts. It means that they denote someth, ng 
which is essentially dliarma. How can they ultimately denote 
the substance? If they do not, the view that all words 
denote the dravya’ falls to the ground. This difficulty is 
removed as follows:— Really, these limiting factors are not 
different from the substance. If they were different, they 
would be nothing at all. They exist as one with the ulti¬ 
mate, the apeidhimat. To be one with it is their essence. 
They are not the essence of the ultimate. The limiting factors 
are so when they are merged in the ultimate. The moment 
they are detached from the ultimate, they would cease to be 
limiting factors. They would acquire a kind of indepen¬ 
dence and themselves become upddhimat .] 

7. The tradition which has come down from the 
ciders is that there is no difference between the real 
and the unreal. The real, when not properly under¬ 
stood, is called the unreal. 

[It might be said that, if, what is an attribute (dliarma) 
now may, in other circumstances, become the dharmin 
and thus become real and eternal, the principle that the 
form is unreal and the substance real would break down. 
To meet this objection, it is pointed out that in monism, the 
real and the unreal are not two totally different things, 
because, if they were, monism itself would be undermined. 
There is only one ultimate reality, which, due to nescience^ 
appears in many forms to different experiencers. The one 

III. 2.8-10 


stands for knowledge and the many for nescience. When 
the one manifests itself as many, it is prapanca, which is 
wonderful as long as one does not probe too much into it.] 

The author now states that it is Brahman which appears 
as differentiated. 

8 The undifferentiated Reality appears to be 
differentiated. There is really no distinction of time 
within it and yet such a distinction is cognised. 

[Thus, it is Brahman which manifests itself now as this 
and now as something else. Therefore, all words ultimately 
denote that. The Reality which is devoid of all differentiation 
appears to have it on account of nescience. Through the 
powers called Dik (Space) and K<11 n (Time), this one Reality 
presents itself as having spatial and temporal sequence.] 

The author now explains how the non-existent appears 
to have existence. 

9. Just as the attributes of the object cannot 
belong to cognition at all and yet that which is not 
identical with ii appears as one with it. 

[According to the Vijnanavadins, the external world 
does not exist and yet objects appear in our mind as being 
external. Consciousness appears to have various forms which 
appear as external. In other words, the unreal appears in 
the real.] 

10. In the same way, the forms of the transforma¬ 
tions do not at all belong to the Reality and yet that 
which is not at all identical with it appears as one 
with it. 



[Another illustration of the same would be this: Pri¬ 
mordial matter, according to the Sankhyas, contains within 
itself, the germs of all later transformations and yet, it is free 
in its own nature, of all these transformations. It consists of 
a balance or equilibrium of the gunas and is, therefore, quite 
different from that state where there is disturbance of the 
equilibrium. But, in actual life, it can be cognised only 
through its transformations which have not the same reality. 
Thus Reality is manifested through the unreal. This principle 
is accepted by the Buddhists as well as the Sankhyas.] 

The author now states how we know that the forms are 
unreal and what is different from them is real. 

That is real which persists till the end. when 
ail the forms disappear. It is eternal, it is expressed 
by the word and it is not different from the ultimate 

[It is stated in the M. Bha: “That is also eternal which 
does not lose its identity (Tad api hi nityevm yasmhnstattvavi 
na vihmiyate (M. Bha on Va I, paspasdhnika ). The forms of 
the different ornaments vary and are impermanent, but the 
gold persists in all of them. Similarly, Brahman is eternal 
and real. The eternality of the universal and other things is 
only relative. Compared to the different individual cows, 
cow-ness is eternal. Compared to cow-ness, horse-ness etc., 
earthness ( prihivitva ) is eternal. Compared to that, the fact 
of being just a thing is eternal, an idea which can be con¬ 
veyed by the pronoun ‘tat’. But, right through these more and 
more comprehensive and yet only relatively permanent things 
persists consciousness and, therefore, that is the real truth. 
And this consciousness is nothing more than pasyanti or 
para vdk, or sabdabrahman. Brahmtm is not different from 
Pabda. Brahman is the vacya of all words which are its 
vdcalca-s. Vacaka and vacya are not different from each 
other. From the way Helaraja explains the Karika, his text 
probably was: tacchabdatattvanna bhidyate .] 

III. 2.12-13 


The author now shows how the Reality is absolutely 
different from all appearances. 

12. It does not exist nor does it not exist; it is not 
one nor is it different; it is not connected nor is it 
separated; it is not transformed nor is it not so. 

[The Reality is beyond all transformations. It cannot, 
therefore, be identical with them. It is beyond all assertions. 
One cannot say that it exists nor that it does not exist. Nor 
can one say that it is one, because the Reality which is free 
from all limiting factors is without 'any inner differentiation 
and, therefore, it does not appear as one. Nor does difference 
add any peculiarity to it, because there is nothing different 
from it and whatever is different from it has no reality. Nor 
can connection or separation be its limiting factor because 
that presupposes the existence of something different from it 
and that cannot be proved. It cannot undergo any real trans¬ 
formation, because one does not accept such a thing; one 
accepts appearance instead ( vivarta ). And yet the whole 
wonderful universe is manifested, so that there is some kind 
of change. Thus it is not possible to make any positive asser¬ 
tion about the Reality.] 

The author now says that everything is identical with it. 

13. It does not exist and it does; it is one and it 
is many; it is connected and it is separated; it is trans¬ 
formed and it is not. 

[And yet it is Brahman which appears as everything else. 
It appears as positive entities and as negative entities. It 
appears as one and as many (as one in the case of universals 
and as many in the case of the individuals. It appears as 
associated with other things and as separated from them. It 
appears as transformed and as not transformed.] 

The author now says that, as everything is Brahman , 
even contradictory usages are resolved in it. 



14. That one Reality is seen as the word, the 
meaning and their relation. It is the seen, the seeing 
the see-er and the fruit of the seeing. 

[The expression, the expressed and their relation are one 
in the ultimate reality. In it, the power of the expression 
(vacaka, sruti&akti) and that of the expressed (vdcya, 
arthasakti) are united. In the manifested state, it appears 
as two branches, the word and the meaning (sruti and ariha). 
Manifestation as knowledge ( jhana ) and the known (jheya) 
separately is nescience. This has been explained in the 
Brahmukdnda. It is the same reality which evolves as the 
see-er and the seen. What is called the seen, or the objective 
world figures in the consciousness. Its very essence is, there¬ 
fore, consciousness. If it were not so, it could not be illumi¬ 
nated. It has also been declared that the see-er, the indi¬ 
vidual soul, limited by nescience, the transmigrator, is really 
Brahman, because, being conscious, cannot be really different 
from the ultimate reality. The mention of the two accessories 
(karaka) namely, the agent and the object ( kartd and karma, 
i.e. the see-er and the seen) is meant to include the other 
accessories also. Thus the whole of the objective world 
(siddha) is included. The word darsana refers to the main 
action and includes all other actions and so the whole of the 
world of processes ( sadhya ) is included. The manifestation 
called action is determined by the power called Time and 
the manifestation called concrete object or embodied object 
(murtti) is determined by the power called Space (Dik). 
Thus, the whole universe, consisting of things and processes 
(murttivivarta and kriydvivarta ) is set forth. The word 
prayojana in the verse stands for the fruit of all actions. Thus, 
all the three things, sadhya, sddhana and phala are included 
and it has been shown that the whole of it is a vivarta of 
Brahman. This has already been explained in Vak. I. 4. 
The ‘seen’ and the ‘seeing’ are separately mentioned because 
usage consists of prakhyd, i.e., cognition or ‘seeing’ and 
uvakhyd, i.e., giving expression to cognition by means of 
words involving sabda, ariha and sambandha which together 

Hi. 2.15-16 


stand for the ‘seen’. All this, of course, is in the state of 
nescience. In reality, all this differentiation does not exist.] 

The author now shows that when all the appearances 
disappear, something unchangeable does remain. 

15. Just as, when forms disappear, it is the gold 
which is the truth ( satya) in the ear-ring etc., in the 
same way, when transformations (like earth etc.) dis¬ 
appear, the primordial substance is the only thing which 
is real. 

[That which persists when all the forms disappear has 
been declared to be real. But somebody might ask whether 
something does persist at all. In answer, it is said that when 
forms such as ear-ring (kundala) disappear, the gold remains. 
Similarly, when forms such as earth disappear, primordial 
substance, that is, Brahman , remains. For this, the authority 
is the written tradition (agama). It would not be right to 
say that all this universe proceeds from something which is 
non-existent and inexpressible. Non-existence cannot pro¬ 
duce existence. Nothing can come out of hare’s horn. 
Consciousness persists in everything. Therefore, everything 
originally came out of it.] 

The author now concludes by saying that all words ex¬ 
press that. 

16. The primordial substance is the expressed 
meaning of all words. The words themselves are not 
different from it. Though not different from one 
another, there is a relation between them as though 
they were different from one another. 

[Thus, what is called substance ( dravya ) is really Brah- 
man, the only persisting reality. It is that which is expressed 
by all words. All usage is based on differences brought about 
by limiting factors. It is based on nescience. All words 

V-III. 10 



therefore, express Brahman, differentiated on the basis of 
limiting factors. Even words like atmd, brahman, tattva ex¬ 
press that primordial substance through some limiting factor 
or other. Because that which is beyond all limiting factor 
(nirupadhi) is also beyond the range of words. In compari¬ 
son with words like ghata, words like atmci are much nearer 
to the ultimate reality. In fact, words themselves are not 
different from the ultimate Reality, in the world, one talks 
as if they were different from one another.] 

The author now gives an illustration to show that the 
manifestations are unreal and only unity is real. 

17-18. Just as, in a dream, the one mind appears 
in contradictory forms, as the self and the non-self, 
friend and foe, the speaker and the spoken and the 
purpose, in the same way, while the ultimate reality 
is unborn, eternal and devoid of inner sequence, we 
see it as having birth and other contradictory attributes. 

[One should not wonder that all the plurality which we 
see before us is being denied and the unity which one does 
not see is being advocated. Because that kind of thing is 
happening all the time. The world which we see in our 
dreams is contradicted in the wakeful state. Similarly, the 
world which we see in the wakeful state does not persist in 
the turiya (the state beyond deep sleep). So that may also 
be looked upon as unreal. That which persists in all sfates 
is the only thing which is real. Consciousness is the only 
thing which so persists. The different states come and go 
and they, like pleasure and pain, do not affect the real nature 
of consciousness. The variety which one sees in a dream is 
confined to a particular experiencer. This experiencer, the 
transmigrator, being essentially consciousness, is Brahman . 
He creates things out of himself, without any other material 
and enjoys them. It is this idea which is conveyed in the 
following verse, quoted in the Vrtti on Vak. I. 127. 

III. 2.17-18 


Pravibhajydtmandtmdnam srstva bhdvdn prthagvidhan. 

Sarvesvcirah sarvamayah svapne bhcktd jiravartate. 

The word bhoktd refers to the creation of the individual self. 
As it is ultimately Brahman , it has the power to create. It 
does not require any external material for the purpose. As 
there is no real difference between the creator and the 
creation, it is said to be vaikalpika . The creation which takes 
place in the wakeful state depends upon external material 
and it is a creation which is common to all experiencers. The 
dream creation is impermanent, whereas the one in the wake¬ 
ful state is relatively permanent. Both are unreal, because 
both are manifestations of nescience. In dream a further veil 
on consciousness, namely sleep, also takes place. On account 
of that, the average man looks upon the dream-creation only 
as unreal. But to those who know the truth, the whole 
universe, that of dream and that of the wakeful state, is un¬ 
real. Only the persisting consciousness is real, without 
sequence and eternal. But to nescience, it appears to have 
sequence. Therefore, it is the eternal sakti of Brahman 
which manifests the unreal world, the perceiver and the 
perceived and creates this world dream. It is the function 
of philosophers to remove this universe which is charming as 
long as we do not reflect.] 



Section 3 


• Unless there is a relation between the word 
its meaning, any word would convey any meaning and 
does not happen. So the relation between the two is 
being explained — 




1. From words which are uttered, the intention 
of the speaker, an external object and the form of the 
word itself are understood. Their relation is fixed. 

[When words are uttered, three things are understood: 
(1) their own form; (2) an object which is a means to fulfil 
a certain purpose; (3) the intention of the speaker. The rela¬ 
tion between these three things is fixed, that is, not made by 
man] ■ Of these, the first is closest to the word ( antaranga ). 
That is why it is understood in any case. Between the own 
form of a word and its meaning, there Ts the relation of the 
expression and the thing expressed (vacyavdcakabhdvaj . 
Between the speaker’s intention and the word, the^e is the 
relation of cause and effect ( kdryakdranabhava) ' As the 
science of grammar is common to all systems (sarvaparsada), 
the latter relation is also spoken of, in order to meet the view 
of those who hold that the meaning^ of a word is only what 
figures in the mind of the speaker.In that view, the relation 
between the intention and the word would naturally be that 
of cause and effect, because the intention would evoke parti¬ 
cular words and not others. Whatever be the relation, there 
is superimposition of the word on the meaning ( adhyasa ). 
By ‘word’, it is chiefly the sentence whicn is kept in mind, 
because, according to grammarians, the individual word is 
only an abstraction (cipoddhara ).] 

The author now says that the own form of a word is a 
meaning in a primary sense. 

III. 3.2 


2. There is sometimes knowledge and sometimes 
doubt in regard to the meaning and the intention of 
the speaker. No mistake is possible about the forms 
of the words which are perceived. 

[The word gciuh and the object of which it is the name, 
both appear to us in the form ayam gauh. That means that 
a word conveys an object (a meaning) and conveys its own 
form in the process. This process of conveying a meaning as 
one with itself is called expression ( abliidhana) . That is how 
the convention is understood. The very purpose of conven¬ 
tion is worldly usage and therefore, it has to follow worldly 
usage. In worldly usage, the word and the meaning are 
identified. The senses are only a means to cognition and do 
not become part of it The sense of vision, for instance, does 
not become part of our perception of an object. Certain 
things like signs enter into the cognition which they cause, 
but remain apart from it. For instance, the sign ‘smoke’ 
causes the inference of fire but remains apart. The word is 
also a means of cognition like the senses and signs, but there 
is this difference that the object whose cognition it causes 
appears as one with it. 

f _ 

One might here object to the last statement by saying, 
rather sarcastically, that the letters of the alphabet are not 
found reflected in the object. In other words, we do not see 
any identity between word and meaning. But this objection 
derives from an imperfect understanding of fundamental doc¬ 
trines. That word is said to be expressive (vacaka) which is 
the function called ‘vdk’ of the power of consciousness. This 
function is called verbal expression (sabdana) . Even in 
silent recitation, it is so called. In its first stage, it is un¬ 
differentiated into word and meaning. It remains in the state 
of the supreme word. Afterwards, it enters into the region 
of the mind, and the breath (pranavrtti) and becomes dif¬ 
ferentiated, into the expressed and the expression ( vdcya and 
vdcaka) . In this state which is called the ‘middle one’ 
(madhyama) , the expressive word (vdcaka), without aban- 



doning the state of the supreme word which consists in un¬ 
differentiated consciousness, refers to form and meaning 
closely bound up and standing in apposition to each other in 
the form gaur ayam arthah. There is~ said to be super¬ 
imposition of one over the other (adhydsa) . There is dif¬ 
ferentiation between the expression and the expressed and 
yet the two are identified. In the next stage, the word, while 
still not giving up its basis in the previous stages, becomes 
audible through articulation in the mouth. The word, being 
thus a manifestation of the supreme word, retains its 
characteristic of identity between word and meaning 
(sdbdana) even after the two have been differentiated. In 
one single operation, a complex meaning involving a qualifier 
and something qualified ( visescina and visesya) is conveyed 
as one with the words which express it. For example, in the 
sentence suklah pato’yam = this white cloth, the cloth as 
qualified by the colour white, is conveyed as one single thing. 
The white colour is not understood as something quite distinct. 
Similarly, in ghafo’yam — this jar, the object jar is under¬ 
stood as qualified by the word jar, its visesana. This process 
of understanding the word and meaning as one is called 
sabdanc^~ ‘Sometimes, the meaning of the word is the form 
of the word itself as in agnim antodattam adhisva — ‘recite 
agni\ with the udatta accent on the last syllable’. Where the 
meaning is an external object, the w-ord and the meaning 
appear as one. They are not cognised separately. The word 
conveys both its own form and the meaning by its expressive 
power ( abhidhd ). Three things communicate themselves 
while communicating other things: Cognition, the word and 
the lamp; Thus, the word conveys its own form and its 
meaning as the expressed sense. It is not that its form merely 
becomes the object of the sense of hearing. Its form and its 
meaning become the contents of the same cognition. They 
are in apposition to each other ( samanadhikarana ). The form 
as the expressed sense is different from the same thing as 
expressive of the meaning. In the latter capacity, the word 
is active (hartrbhuwihn uparudhri), it is consciousness 
(paramarsa), it is the supreme word (para vale), it is light 

III. 3.3 


(pvakdsa) and, therefore, it is different from tlie word as the 
expressed sense (vcicya) , as the thing one is conscious of 
(paramrsyamana ). When the hearer hears the words of 
others, he understands their meaning as identical with tnem 
and infers the meaning of the speaker also as identical with 
the words. Thus, there is superimposition, not only between 
the word and the meaning as understood by the hearer, but 
also as expressed by the speaker. "So, from a word like gliata, 
we understand three things all mixed up: the word, the 
meaning and the intention of the speaker. That is why the 
question gaurityatra kali sabdali, has been asked in ihe 
Malidblidsya in order to isolate the word from this complex 
cognition. Even the cognition of a new-born child is mixed 
up with the word. Thus, whether the relation is looked upon 
as fitness or causality, superimposition ( adhydsa ) is involved 
in both. 

Thus, due to the function called sabdana , which is the 
very life of what is called illumination ( prakdsa ), both the 
word and the moaning become the contents of this cognition 
and both are in apposition to each other and identity is super¬ 
imposed on them ( abhedddhydsa ). In 4 Pasyanti , where 
there is no differentiation, the question of differentiation does 
not arise at all. v The word remains the very life of meaning 
even in the stage of verbal usage (vyavahdra) and that is 
why the whole universe is said to be a manifestation (i rivarta) 
of the word. 

H The form of the word is its first expressed sense because 
it is part of the process ( antaranga ), it is unavoidable (ahoya) 
and it is distinctive (asddharana ). Here there is no depen¬ 
dence on convention. It may be looked upon as a universal 
or otherwise, as explained in the Brahmakdndci.'] 

The author now says what the indication is of the existence 
of the fixed ( samavasthita ) relation between -'word and 

3. That this word is expressive of that meaning, 
that this meaning is the expressed meaning of that 




word, such a relation between word and meaning is 
understood from the genitive suffix. That is why they 
are understood as identical. 

[That the relation between the word and the meaning is a 
natural one and not a mere matter of convention is inferred 
by us in the following way. We use such expressions as: 
This word is expressive of that meaning; that meaning is the 
expressed sense of this word. Here the use of the ‘of’ 
(sasthi vibliukti) can be explained only on the assumption 
that there is a natural relation between the word and the 
meaning. We never connect two such totally unconnected 
thmgs as the jar and the cloth by means of the genitive suffix. 
Because of this natural connection it is that we identify the 
two in such expressions as ayam gauh where the word gauh 
and its meaning are identified.] 

4. There is no verbal element (besides the geni¬ 
tive suffix) which denotes this relation in its essential 
property. Being extremely dependent, its own form is 
never cognised. 

[There is no word which directly denotes relation. 
Ordinarily we observe things and their properties and use 
appropriate words to define or describe them. We cannot do 
that in the case of relation, because we cannot observe it 
objectively (idamtaya svarupanavadharandt ). Qualities like 
‘white’ reside in some substance or other and yet words like 
sukla sometimes express the whiteness in isolation. There is 
no word which can express this relation in isolation because 
it is always dependent on something ( nityaparatantra ). It is 
nothing more than a certain mutual requirement of two things 
(sambandhinoh parasparaunmukhyasvabhdvah ). It is a mere 
attribute ( dharma ). It has, therefore, no word of its own 
to express it. It has to be inferred from its effect ( kdryailcaga- 
myall ).] 

in. o.5-7 


5. Where there is service rendered, there an attri¬ 
bute (that is, relation) is understood. It is the power 
of powers, it is the attribute of attributes. 

[Where one thing renders service to another, one can 
infer a relation between the two. Two unrelated things can¬ 
not render service to each other. Relation is not the same as 
power or capacity, because it is the former which regulates 
the production of a particular effect by power or capacity. 
Power resides in a particular thing and it is relation which 
regulates it. That qualities reside in substances is regulated 
by relation.] 

The author now says why relation is being spoken of, 
considering that others have already spoken about relations 
such as conjunction and inherence. 

6. Conjunction and inherence are called relations 
because they have the attribute thereof. They have 
definite conditions the purpose of which is to render 

[Conjunction and inherence do possess the chief 
characteristic of relation, namely, dependence. The applica¬ 
tion of the word relation to them is really a case of secondary 
usage. Primarily, it is not they which constitute relation. 
When they themselves are conditioned by other things, the 
latter become dependent (paratantra ) and they become 
svatanlra and as dependence is the chief characteristic of 
relation, they cease to be so. They can have their own updclhi 
which render them service. When the relation between them 
and their upadhi is regulated by a relation, how can they 
be identified with something which is purely dependance?] 

The limiting factor of conjunction is now being stated. 

7. There is only one particular state in which 
the conjunction of one thing with another takes place 
and that is indicated by the effect. 

v-m. li 


[When the juxtaposition of parts results in their contact, 
the whole is produced. Thus, in order to explain the produc¬ 
tion of the whole, contact is taken to be the non-inherent 
cause. The fact of being confined to substances and connec¬ 
tion with Being as a quality, this is the limiting factor 
(upddhi) of conjunction. It is relation which determines the 
service which it renders. That is why conjunction is not the 
same thing as relation. The latter can exist in substance, 
quality, action, universal etc., but conjunction is confined to 
substances. Conjunction has the effect of producing a posi¬ 
tive concrete thing whereas relation only brings two things 

The difference between relation and inherence is now 
being pointed out. 

8-11. It has been traditionally handed down from 
the elders that Relation is beyond the characteristics 
of all things (padartha ), that it is made known by 
everything ( sarvalaksana ) and that it favours that 
power called inherence ( samavdya ) by virtue of which 
the production of things (previously non-existent is 
restricted to a particular (cause) by which the effect 
is never separated from such a cause, there being no 
difference (from conjunction) in the matter of juxta¬ 
position, by which the essence of something is cognised 
as that of something else, by which neither distinction 
nor identity of something is perceived, which renders 
assistance to other powers, which is beyond identity 
and difference and which is established differently from 
other things. 

[According to Satkaryavada, the effect already exists in 
the cause. Hence, it is the cause which becomes the effect. 
So particular effects can be found in particular causes only. 
According to the ' cisatkaryavada’ also, a particular effect like 
cloth can exist only in the threads and so on. Thus, the effect 

III. 3.8-11 


is intimately associated with its cause, it is not separated from 
it even though they are different from each other. This is 
due to inherence ( savtavdya ) and this is where it differs 
from samyoga, though the two resemble each other in one 
respect, namely, that they bring two things together. It is 
due to inherence that cause and effect; part and whole, the 
universal and the particular, the quality and substance, 
though different from each other, appear to be the same. 
They only appear to be the same, because there is difference 
between the two in each pair. It is because of this that 
thinkers hold different views on this matter. Some think 
that the parts and the whole are the same while others think 
that they are not. What all this means is that the fact of 
being found together, of being intimately united (ayutasiddhi ) 
is the limiting factor of inherence. It causes the cognition 
of a particular substratum (ihapratyaya ) between things 
which are intimately united ( ayutasiddha ) and not between 
others. It is Relation which determines this characteristic of 
inherence. Inherence does not appear to be different from 
the two things which are united by it, just as the two things 
themselves do not appear to be different from each other. 
All of them appear to be mixed up ( sammurchitarupatvend - 
vabhdsdt) . Inherence is inferred to exist from its effect, 
namely, the cognition of a particular substratum (ihapratyaya ). 
Inherence is here called a power. As it renders service to 
things, it cannot be put in the same category as them. As it 
is a power, one cannot say whether it is identical with that 
of which it is the power or not. It is not a thing (paddrtha), 
it has to be put in a different category. That is why it is 
said to be something which is established differently (anya- 
thaiva vyavasthitdm) . Inherence renders service to the 
powers called sddhana by making them reside in their 
substrata. It is such an inherence which is helped by Rela¬ 
tion which (1) is not a thing because it lacks that indepen¬ 
dence which exists in all things, (2) is made known by 
everything which needs help, and not by fixed things as 
inherence is, (3) confines inherence to things which are inti¬ 
mately united, (4) gives to inherence the name of relation 



because of the common characteristic called dependence. 
Relation which does all these things must necessarily be 
different from the p adartha-s. That is why, in this sdstra, it 
is not merely conjunction and inherence which are relations.] 

The author now explains this point further. 

12. Others accept relation everywhere as a separate 
category. But that cannot explain the relation between 
word and meaning. 

[The Vaisesikas have put conjunction among the quali¬ 
ties, that is, it comes under one of their categories. They 
have made inherence into a separate category, even though 
it is a kind of power and, therefore, resides in something else. 
Tiie fixed relation between word and meaning which finds 
expression in such statements as: ‘this is the meaning of this 
word' cannot come under inherence. Therefore, there is in¬ 
completeness in the Vaisesika conception of the categories.] 

13. Through inherence, words convey their 
substratum and their universal. Through inherence in 
the same thing, they convey the qualities which exist 
in their substratum. 

[Words like kha, dkdsa, gagana convey their meaning 
through the relation called inherence, because they are words 
and all words are qualities residing in dkdsa by inherence. 
All words convey the particular universal existing in them 
through inherence. As for the special qualities like mahattva 
which exist in dkdsa, words convey them through inherence 
in the same thing ( ekdrthasamavaya ), because both words 
and these special qualities exist in the same thing, namely, 
dkdsa, by inherence.] 

14. The universals ‘substance-ness and Being and 
conjunction are linked to their own and other substrata. 
So are qualities like duality etc. 

III. 3.15-16 


[A word like drcivyatva conveys substanceness, the lower 
universal and Joeing, the higher universal through the 1 ela¬ 
tion of inherence in the same thing. They exist in dkasa , 
the substratum of the word drcivyatva as well as in earth etc. 
Thus the universals of words exist in their own substratum 
as well as elsewhere through inherence.] 

15. Some objects are in conjunction with the 
substratum of their words, some, on the other hand, 
inhere in them; others still inhere in the things which 
inhere in those which are in conjunction. 

[Scibda, being a quality, inheres in dkdsa which is, there¬ 
fore, its substratum. Objects like chariot, jar etc., are con¬ 
nected with dkdsa by the relation of conjunction. These ob¬ 
jects are conveyed by their words through saviavetascimyoga, 
conjunction with the intimately united. The qualities and 
movements inhering in objects like a jar are conveyed by 
their words through svdsrayasamyuktasamavdya } inherence in 
what is in conjunction with one’s own substratum ( svdsra - 
yena akdsena samyukto yo ghatah tatra rupasya samavdya- 
sambadhena vrltili) . The quality of being colour (juputva) 
which exists in the colour which exists in the jar is conveyed 
by the word rupatva by means of svdsrayasainyaktcisaviaveta- 
gamavdya , inherence in what is inherent in what is in con¬ 
junction with one’s substratum.] 

16. A word conveys an all-pervasive object which 
is in contact with what is in contact with its substratum. 
But no other relation is seen for inherence. 

[Space, Time and Soul are all-pervasive. They cannot, 
therefore, come into any new and direct contact with dkdsa 
which is also all-pervasive. They are conveyed by their words 
(. Dik, kdla, dtmd ) by the relation of sv&srayasamyiikta- 
samyoga in the following way: Earth etc. are in contact with 
dkdsa . Space, Time and Soul are in contact with earth etc. 
The qualities which inhere in all-pervasive objects are con- 



veyed by their words by svdsrayasamyuktasamavdya. The 
universal which inheres in these qualities is conveyed by 
svdsrayasamyuktasaviavetasamavdya. Thus, how substance, 
quality, action and universal are conveyed by words has been 
shown. The word visesa conveys the particularity (visesa) 
inherent in atoms by samyuktasamavaya. It conveys the 
particularity inherent in akdsa by inherence in the same 
thing. It conveys the particularity inherent in Dik, kdla and 
atmd by svdsrayasamyuktasamavaya. Thus the way in which 
words convey five out of the six categories accepted by the 
Vaisesikas has been explained. Nothing has been said about 
inherence, also one of their categories. It is not a substance 
and so cannot be related to anything by conjunction. Nor can 
there be another inherence. Nor can one argue as follows: — 
‘‘The word samavaya, being a word, inheres in dkdsa. Jar 
etc. are in conjunction with akdsa which is all-pervasive. 
Inherence exists in jar etc. Therefore, the word sam,avd.ya 
conveys its meaning by svdsrayasamyuktasambandha , relation 
with what is in conjunction with one’s own substratum”. 
Inherence can exist in jar etc. only by inherence and there 
cannot be a second inherence.] 

17. As there is no special relation, there would he 
no fixity (in the conveying of meaning by words). 
Therefore, relation (consisting of conjunction or 
inherence) is not applicable to words and meanings. 

[If it is maintained that there is no special relation by 
which the word samavaya conveys its meaning, that would 
be impossible. There must be some relation. Otherwise, the 
word would become meaningless. If the word samavaya 
conveys that category without any special relation, any word 
would convey any meaning, because the absence of a special 
relation is not a bar to the conveying of a meaning. Saviban- 
dhasyavisistatvdt may also mean: because the relation is the 
same as in other cases. That is, if the word akdsa conveys 
its substratum (ether) through inherence, the word ghata. also 
should convey ether through inherence, because it is also a 

iii. 3.18 


word and, therefore, a quality of tiktisa. The word gluita 
would also mean pata, because conjunction with the substra¬ 
tum aktisa exists in the case of pata also. Thus, there would 
be both narrowness and excessive width (avyapli and ati- 
vytipti ) in the relation between word and meaning. There¬ 
fore, the attempt to explain the way in which words convey 
their meaning by having recourse to conjunction and 
inherence is a failure.] 

The author now hints at a possible justification of the 
Vaisesika view on the basis of mere logic (praudh ivada ). 

18. Just as the Self, while it is connected (with 
everything) has the relation of ownership in regard to 
certain things only through adrsta, in the same way, 
elsewhere also the same thing holds good even though 
the relation is the same. 

[Just as the all-pervasive Soul which is in contact with 
everything feels a sense of ownership only in regard to the 
body, in the same way, even though all words, being words, 
inhere in tiktisa , each one through adrsta , means some parti¬ 
cular thing only. Thus there is no excessive width. One 
might argue thus from the Vaisesika point of view, but it is 
not right. Even though the Self is all-pervasive and, there¬ 
fore, in contact with everything, yet everything does not 
belong to the Self nor does the same thing belong to all the 
selves, because what belongs to which self is regulated by 
adrsta. In the same way, all things are not expressed by all 
words. Which word expresses which thing is regulated by 
usage and convention. 'The relation between word and mean¬ 
ing is fitness ( yogyatti ). There is, therefore, no use of think¬ 
ing about conjunction and inherence in connection with word 
and meaning.] 

The author really holds that inherence is not the ex¬ 
pressed meaning of any word. 



19. Neither the speaker nor the hearer can reach 
through words that juxtaposition (prcipti) called 
inherence which is beyond the attributes which things 
expressed have. 

[All things like a cow which are expressed by words are 
characterised by some properties and are mixed up with 
them. These two things are their limiting factor. But for 
them, they could not be expressed by words. Inherence has 
no quality or universal as its limiting factor, according to 
the Vaisesikas. Thus, it is beyond the range of what is cha¬ 
racteristic of all that is expressible ( vdcya ). The capacity 
to be referred to by a pronoun as ‘this’ or ‘that’ is the sign 
that something can be expressed by words. This is also 
absent in the case of inherence. It cannot be understood 
even from the word samavaya by the speaker or by the 
hearer. There is no word which can touch its essence. The 
word samavaya is not such a one, because it is also used to 
denote an assembly of men. It is only among the Vaisesikas 
that it is used to denote this category]. 

Now the Naiyayika indulges in some casuistry. 

20. If what is said to be inexpressible is expressed 
as inexpressible, it would thereby become expressible. 

[The Naiyayika points out that to say that inherence is 
inexpressible ( avacya) is a contradiction in terms because 
it comes within the range of the expression ‘inexpressible’ 
( avacya ) ]. 

21. If it is said to be not capable of being expressed 
even by the word 'inexpressible’ the condition that is 
meant to be conveyed would not be understood. 

[If it is maintained that it is not expressed even by 
the word ‘inexpressible’, then what is intended to be con¬ 
veyed by all this is not understood, namely, that inherence 
does not come within the range of words]. 

III. 3.22-23 


22. If what is meant is that it is inexpressible in 
some aspect or other or in all aspects, even then those 
words would not deny that particular condition. 

[If what is meant is that ‘inherence’ is not expressible 
in its nature as dependence ( pdratantrya ) it is as good as 
saying that it is not expressible at all. There cannot be a 
negation of an attribute ( dharma ) without there being a 
negation of that which has the attribute (dharmin) . There¬ 
fore, even if only its dependence aspect is declared to be in¬ 
expressible, it itself becomes inexpressible. Even to deny 
any particular aspect of a thing, it has to become the object 
of verbal knowledge and if it becomes the object of verbal 
knowledge, it has become ‘vacya\ How can it then be said 
to be avacya’?~\ 

The author now answers the casuistry employed by the 
Naiyayika by giving an example. 

23. When knowledge in the form of doubt has 
assumed the form of sesa (meant for something else), 
it cannot become the object of another doubt without 
losing its original form. 

[When the particular form of a thing like a post is not 
perceived, a doubt arises: is this a post or a human being? 
This cognition in the form of a doubt is chiefly concerned 
with finding out the true nature of the object which figures 
in it. Just at that time, another doubt, doubting this very 
doubt, cannot arise. The first doubt is concerned with find¬ 
ing out the nature of the object. Just at that time, it can¬ 
not become the object of another doubt. If it does, it would 
lose its former condition of being a doubt. As a doubt, it 
was visayin and as the object of another doubt, it would 
become visaya and one and the same thing cannot become 
visaya and visctyin at the same time.] 

The author now gives another example. t 

v-m. 12 


24. When, in a decisive cognition, decision as such 
takes place, then the cognition retains its characteristic. 

[When a decisive cognition is engaged in clearly deter¬ 
mining the nature of the object, it cannot become the object 
of another cognition. If it did, it would lose its proper nature 
and become the ‘object’ of another cognition. To be en¬ 
gaged in determining the nature of the external object is 
the nature of cognition. When it is so engaged, it cannot 
be the ‘known’ (j neya ). In the same way, the nature of a 
word is to convey a meaning, to bring the object within its 
range; to say abdut something that it is ‘avacya’ which is 
a word and to say at the same time that the thing is inex¬ 
pressible is a contradiction in terms.] 

25. The sentence ‘all that I am saying is wrong’ 
is not literally meant. If what it says is wrong, the 
point in question would not be conveyed. 

[When one says: ‘all that I am saying is wrong’, the 
incorrectness should not be applied to that very sentence. 
Otherwise, the incorrectness of what the person had been 
saying before — the point which is sought to be conveyed 
— would not be conveyed. On the other hand, if that sen¬ 
tence is held to be expressive even though it is incorrect, 
the same could be said of the previous sentences and that 
would mean that what was meant to be conveyed would 
not be conveyed.] 

26. What is expressive cannot at the same time 
be the expressed. What conveys something else can¬ 
not at the same time be conveyed by something else. 

[What is engaged in conveying something else cannot 
turn back and become the object of the expressive activity 
of something else. What is acting as the agent cannot at 
that very moment become the object. What is expressive 

III. 3.27 


cannot at that very moment be the expressed. The follow¬ 
ing syllogism expresses the idea well. 

Yat pratipadakam na tat pratipddyam , 

Yalhd sainsayajnanam nirnayajndnam vd 

pratipadakam r.dbhidlidnam, 
ato naitat pratipddyam. 

“What conveys something else cannot itself be conveyed 

by something else. 

As, for example, a doubt or a decisive cognition. 

The word conveys something else, 

And so it cannot itself be conveyed 

(by something else).] 

27. The statement: “a mere assertion does not 
prove the point” does not apply to that statement .itself. 
Similarly, the property (incorrectness) is not under¬ 
stood as applicable to this sentence. 

[It is true that incorrectness cannot be attributed to the 
meaning of the sentence sarvayi mithyd bravimi without its 
giving up its expressivity (vdcakatva) and becoming the 
expressed ( vdicya ). But incorrectness may be attributed to 
it in another way. After all, it is also one of the sentences 
uttered by the person in question and if all that the person 
says is wrong, this sentence would also be wrong. To meet 
this objection, another illustration is given. Sddhyanirdesah 
pratijnd sddhandngam na bhiavati is a statement of the 
Bauddhas which means: “A statement of the thing to be 
proved is what is called assertion and it cannot become part 
of the proof.” This does not apply to this very statement. 
Otherwise, it cannot convey what is intended to be convey- . 
ed. Therefore, the statement that assertion is not part of 
the proof does not apply to this assertion. It applies to 
others. In the same way, the incorrectness mentioned in the 
sentence; c sarvam mithyd bravimi’ does not apply to that- 



sentence itself. All this reasoning applies to the word avcicya 
as applied to samavciya.’] 

28. Since one function has not another function, 
therefore one should not point out anywhere contradic¬ 
tion or regressus ad infinitum. 

[‘Anywhere’ means not only in the matter of being the 
expressed meaning, but also in regard to relation. Or it 
might be understood as relating to what was said in verses 25 
and 27.] 

The author now proceeds to explain relation as accepted 
in his own circle. 

29. Just as the senses have an eternal fitness to 
perceive their objects, in the same way, the relation of 
words with their meanings is an eternal fitness in regard 
to them. 

[Because we understand a meaning as soon as we hear 
words, a natural capacity in the words to convey the mean¬ 
ing must be presumed. This natural capacity is made known 
through convention ( sanketa ). This capacity is as natural 
as that of the eye to perceive form and colour. This capa¬ 
city has to be understood by us on the basis of the meaning 
which we understand from words. As words and their 
meanings are understood as one, there is a kind of mutual 
superimposition ( adhycisa ): the word is understood as the 
meaning and the meaning is understood as the word. Adhycisa 
or superimposition is the main relation between word and 
meaning. The senses are causes (karaUa) in the process 
of the production of the cognition of objects. The word, 
on the other hand, makes known (jnapaka) the meaning. 
Being so, it must itself be known before it can cause the 
meaning to be known. The resemblance between the senses 
and the word is that their power or fitness is natural, not 
due to human effort.] 

III. 3.30-31 


The author now points out that the meaning is not relat¬ 
ed to the corrupt form of a word. 

30. The corrupt form, according to some, is expres¬ 
sive only through the inference (of its correct form). 
Even if its expressivity is equal (to that of the correct 
form) the sdstra makes a restriction in regard to usage 
keeping merit and demerit in view. 

[This capacity belongs to the correct forms of words 
and not to their corrupt forms. People who know do not 
understand the meaning directly from the corrupt forms of 
words which have, therefore, no connection with meaning. 
They are not expressive (vdcaka). They cause us to infer 
the correct words from which the meaning is understood. 
But corrupt forms of words have become current and igno¬ 
rant people understand meanings directly from them. But 
merit comes only from the use of the correct forms. The 
correct form of a word is only one while its corruptions may 
be many. Just as Brahman is One and stands for vidyd 
and the multiplicity of the world for avidya, in the same 
way, the correct word stands for vidyd and its corruptions 
for avidya.] 

31. The relation which the word relation expresses 
is this capacity (fitness) and it is through that that the 
word ycgyatd itself conveys the meaning of fitness. 
This capacity is made known through convention just 
as the relationship between parent and child. 

[It was pointed out before that if inherence (samavdya) 
is postulated as the relation between word and meaning, 
there would be difficulty of having to postulate another in¬ 
herence for the word samavdya itself. This difficulty would 
not arise if fitness is accepted as the relation between word 
and meaning. The word e saii'ibav.dha would denote the re¬ 
lation consisting of fitness through its fitness to do so. Simi- 



larly, tliG woid samavaya would. dGnote inherence through 
its fitness. Even the word yogyata would denote its mean¬ 
ing thiough its fitness. This eternal fitness of words is made 
known through usage or convention, the uninterrupted prac¬ 
tice of elders (vrcldhavyavaharaparampard) . Mere conven¬ 
tion cannot give words that fitness if they did not already 
have it before, just as usage can only make known the rela¬ 
tion of parent and child if it was already there. The sages 
also do no more than make known which word means what 
on the basis of this eternal fitness.] 

The relation of cause and effect is now being stated. 

32. The word is the cause of the meaning which is 
indeed produced by it. Similarly, it is from the mean¬ 
ing which is in the mind of the speaker) that the word 
(arises and) is heard (by the listener). 

[Of the meaning which figures in the mind of the liste¬ 
ner, the word is the cause. That meaning presents the thing 
as something external. The meaning which was in the mind 
of the speaker before he spoke is the cause of the word 
which he subsequently utters and which the listener hears. 
Thus the word and the meaning are causes of each other. 
They are conceived and perceived as mixed up, as identi¬ 
fied with each other by both speaker and hearer. The 
meaning which the speaker intends to convey is already unit¬ 
ed with the word. It appears as inner speech. Even the 
cognition of animals is mixed with this eternal word (and- 
disabdabhdvand) . The word which the hearer hears is 
mixed up with the form of the meaning and the meaning 
which he ultimately understands from it is mixed up with 
the form of the word. What is called verbal usage ( vyava- 
lidira) is the exchange of ideas between the speaker and 
the hearer. What the word does is to transfer to the hearer 
the meaning which w'as in the speaker’s mind and which 
was already mixed up with the word.] 

If the meaning of the word is mental, how does one say: 
‘he eats rice’? 

III. 3.53-34 


33. The ideas of meal etc. which one considers to 
be impossible on the view that the meaning of a word 
is mental becomes possible when, after one mental 
meaning is understood, another one comes into 

[According to the Viyhanavddins, the external world does 
not exist. Both action and its accessories are mental. There¬ 
fore, the relation of accessory ( sadliana ) and action ( sadhya ) 
is also mental. On this view, the ideas of meal, eater and 
their relation which, at first seemed impossible, become pos¬ 
sible. When? When, after the idea of cooked rice has arisen 
in the mind, another one, that of its being eaten, also arises 
in the mind. The meaning of one word in a sentence is neces- 
sarily incomplete. It becomes complete only when the 
meanings of other words are also understood. According to 
those Buddhists who accept the external world, the meaning 
of words is something mental.] 

It is now stated that the meaning of a word has eter- 
nality in the sense of continuity. 

34(a.b.) Even in impermanent tilings, there is 
eternality in the form of being the expressed meaning 
of woi’ds. 

[From words, one understands always a meaning in 
which the form of something or other figures. That thing 
may or may not exist outside the mind but as the meanig 
expressed by a word, it is eternal. This eternality is 

It is now stated that it has eternality in the sense of 
unchangeability also (kutasthanityata ). 

34(c.d.) What is called impermanence is nothing 
more than the power which is not different from the 
eternal reality. 



[It is tile power of Brahman, the ultimate reality, which 
causes things to appear as non-existent in the past or in the 
future. This power is not different from the Ultimate Rea¬ 
lity. What is called impermanence or non-existence is only 
an appearance. Every word denotes this ultimate Reality 
through some limiting factor ( wpadhi ) or other, as explain¬ 
ed in the section on Substance. As every word ultimately 
denotes Brahman who is unchangeable ( khtastho.nitya .), 
therefore, the relation between word and meaning is also 

It is now stated that, according to some, even meaning¬ 
less words produce results. 

35. (Even meaningless) words have an influence 
over objects leading to visible or invisible results. How 
is that possible unless there was a (natural) relation 
(between these words and the results)? 

[Some people maintain that another proof that there is 
a nauual xelation between word and meaning is that words 
uttered by Sabaras, even though meaningless, cure cases of 
poisoning. Similarly, magical syllables (bijdksara) known in 
the different sects (though meaningless) produce invisible 
results. All this shows that between words and things there 
is a natural relation.] 

But this is not right. 

36. One does not invariably see such influence on 
objects conveyed by words. Nor is such a relation 
meant here. 

[The meaning understood from words like ghata does not 
produce the same result as the above mentioned words but 
that does not mean that there is no natural relation between 
words like ghata and their meanings.] 

III. 3.37-38 


37. The fact of particular words being the cause 
of the understanding of particular meanings can be 
explained only if there is a relation between the two. 
So it is presumed that there is such a relation. 

[It is better to look at the whole thing as Grammarians 
do. In order to explain the fact that we understand a parti¬ 
cular meaning from a particular word, we assume that there 
is a relation between them, a kind of fitness similar to 
what the senses have in regard to their objects. It is not 
that we infer the existence of such a relation on the basis 
of some visible or invisible result which words might pro¬ 
duce. Such a result may be produced merely by the sequ¬ 
ence of the phonemes of the words through the influence of 
a great personality. It is not due to their capacity to convey 

It is now stated that the relation cannot be mere 

38. Whether the expressed meaning of words is 
eternal or transitory, no human being can establish a 
relation (between words and meanings) with the help 
of words whose relation with their meanings has not 
yet been established. 

[According to the view that the meaning of a word like 
the universal is eternal, nobody can establish a relation 
between word and meaning for the first time. According 
to the view that the things denoted by words are imperma¬ 
nent and that words are also impermanent, the position 
would be the same. The person who would like to establish 
a relation between particular words and particular meanings 
cannot, of course, see all things with his own eyes. Even if 
he sees them through their universals, many universals may 
exist in the same thing and it would be difficult to choose 
between them. Nor can inference be of much help because 
the thing inferred can only be understood as mixed up with 
V-I1I. 13 



words and, therefore, it cannot be of much use in first esta¬ 
blishing a relation between the two. By a process of elimi¬ 
nation, he has to rely on words to do the job. He would have 
to determine things through words and then establish the re¬ 
lation. But the words used for determining things also de¬ 
pend upon convention ( sanketa ) and, therefore, other such 
words would have to be brought in and thus there would be 
regressus ad infinitum . Nor can the nature of things be de¬ 
termined with the help of words the convention of which has 
not been established. Thus one has ultimately to depend on 
words which have a natural and eternal capacity to denote 
their meanings.] 

The eternality of the relation between 
ing is now explained in another way. 

word and mean- 

39. In verbal usage, there is another Being, a 
secondary one, which presents the real nature of things 
in all circumstances. 

[When words convey objects the things so conveyed 
have a Being distinct from their external Being. It consists 
in their figuring in the mind. Such a Being is called aupa- 
carifci’, to distinguish it from Being outside the mind, in the 
external world. Through this Being, things are presented as 
past or as yet to come. That is how words like past (atlra) 
and future ( anagata ) also convey a meaning. When words 
present things as standing in the relation of visesana and 
visesya, it is this Being in which they participate. Things in 
the world exist in a mixed-up state, not visesana separately 
and visesya separately, but words present them separately. 
This separation participates in this Being. The objects de¬ 
noted by words are conceived in the mind and the mind can 
conceive objects which have no external Being and words 
convey them as they are conceived by the mind. That is 
why words like alatacakra and sasavisana have a meaning. 
They convey something conceived by the mind but having 
no external being. Thus words are never without a relation, 
with their meanings. Till something figures in the mind, its 


III. 3.40-41 


existence or Being is a matter of faith only. In verbal usage, 
it is this secondary Being which plays the main part. Things 
come to be because of this Being. They do not assume their 
proper form till they are mixed up with words. When all 
usage can be explained in terms of this Being, if one still wants 
to think of some other kind of Being, consisting of fulfilling 
practical purposes ( arthakriydkarana ), let one do so. But 
such a Being cannot enter into verbal usage. It is another 
matter if such a Being is also reflected in this one. That is 
what is now going to be shown by means of an example.] 

Now follows the example. 

40-41. Just as crystal etc., in association with 
objects of different colours, seem to become one with 
them on account of their power (of assuming different 
forms) in the same way, the word, taking its stand at 
first on this secondary Being, becomes related to various 
properties, contradictory and non-contradictory. 

[The external object has its fixed form, it cannot co¬ 
exist with both existence and non-existence and, therefore, if 
that is the meaning of the word, it cannot enter into relation 
at the same time with the meaning of the words c asti’ and 
nasti. Therefore, the meanings of words have this secondary 
Being. The meaning of a word is capable of entering into 
association with the meanings of all other words. It is like 
the crystal, or glass or the sky. It is sometimes associated 
with the meaning of ‘asti’ and sometimes with that of nasti. 
An object like crystal is very clear and when in contact with, 
say, a blue object, does not give up its own clarity, but ap¬ 
pears to be blue. Similaily, this secondary Being, when link¬ 
ed with such ideas as negation, becomes coloured by it, with¬ 
out giving up its own form. It appears to be identical with 
negation. The clarity of crystal consists in its being able 
to assume various forms and colours. That is its power. The 
power of this secondary Being also consists in its being able 
to assume the forms of all particulars. The word ( sabda ) is 
so called because the object is reflected in it (artliah sabdyate 



iti krtvd ). The word has its footing in this secondary Being 
in the beginning (purvam ), that is, when the meaning of 
the individual word is understood. At the time of the 
understanding of the meaning of the sentence, the meaning of 
the individual word which has its basis in this secon¬ 
dary Being, enters into relation with opposites and non¬ 
opposites. Positive objects are the non-opposites. They 
are favourable to this secondary Being, just as jasmine 
flowers are to crystal. That is why this Being assumes the 
form of these non-opposites in their presence. External Being 
is not opposed to it. In fact, it is very near to it. The oppo¬ 
sites are the negative entities. Being negative in nature, they 
affect the nature of this secondary Being, just as a japci Rower 
which is red affects the colour of crystal. It is due to this 
secondary Being that external objects which are combinations 
of substance and quality are presented by words separately 
as substance and quality.] 

Association with opposites is now being shown. 

42. Similarly, negation relates to the objects of 
negation, resorted to through secondary usage, in order 
that negation may be possible. 

[It is only on the view that the object which figures in 
the mind is capable of assuming different forms that such ex¬ 
pressions as non-brahmin ( abrdhmana ) can be explained, that 
the use of the negative particle in them can be accounted for. 
If words denote things which have an external Being, how 
can we say abrdhmana? In the person to whom the word is 
applied, there is no brahmana-hood in reality. The word 
brdhmana cannot, therefore, be applied to him. But if it is 
applied, the negative particle cannot be joined to it. But on 
the view that the meaning of words has a secondary Being, 
which is common to positive and negative objects, the ex¬ 
pression abrdhmana can be explained. The word brdhmana 
would denote the quality of being a brahmin attributed, say, 
to a Ksattriya on the basis of resemblance. This attribution 
is thus the object which is going to be negatived. It has come 
in through wpacara, secondary usage, which supplies the thing 

III. 3.43-44 


to be negatived, without which there cannot be negation. So 
what happens in cibrahvicfliti is this — The negative paiticle 
denies real brahmanahood attributed to somebody either 
through mistake or resemblance.] 

See M. Bha on P. 2.2.6 and the Pradfpa' thereon. The 
difficulty which arises if words denote things having external 
reality only is now pointed out. 


43. Birth means attaining one’s own form, what 
already exists attains what is to be attained. If it 
(already) exists, why should it be born and if it does 
not, how can it be born. 

Remarks, saita labhyam ca labhyate is a doubtful text. 
sattd should probably be labdha. labhyate may be an error 
for labhasca . 

[If words denote things having external Being, expres¬ 
sions like ahkuro jay ate would be impossible. Jay ate stands, 
for birth. Birth means the assumption of its form by some- 
[hing. There is the thing which assumes its form, the form 
which it assumes and the act of assumption itself. All the 
three exist. That is why one cannot speak about a thing 
being born, as in ahkuro jay ate. That which already exists 
cannot be born. If it does not exist who is the agent denoted 
by the suffix of the verb expressive of the act of being bom? 
If it already exists, how can it be born? The meaning of 
the root also becomes impossible.] 

Another illustration of the same difficulty is given. 

44. If the goer exists and if there is a destination, 
the going can take place. If the one who is born is 
like the go-er, he cannot be born at all. If he is not, 
there cannot be birth at all. 

[The act of going is like that of being born. Only that 
which already exists can go, provided that there is a e -t 
nation as distinct from that which goes. Bui if one \\ io ta ves 



birth already exists like the one who goes, he cannot be born 
at all. Secondly, as birth means ‘attaining one’s self’, there 
is no object as distinct from the agent as in the case of going. 
These are the difficulties if words are taken as denoting things 
having external Being.] 

It is now stated that the difficulties are removed on the 
basis of secondary Being. 

45-46. By conceiving it as agent in order that the 
name (sprout) may be given to it, by further conceiving 
it as the object and by postulating the action (of being 
born) based on it there results the secondary Being. 
The suffixes ‘la’ etc. (the verbal suffixes) can be applied 
to it. But there is no primary Being here because it 
would conflict with the idea of birth. 

[The secondary ‘Being’ is capable of assuming all forms. 
What already exists in its finished form cannot be said to be 
born, nor what is totally non-existent. About something 
which neither is nor is not, which participates in the former 
and later condition and is about to attain a Being, one can 
say that it is being born. When one says that ‘the sprout 
comes out’, what one means is that when the causes are about 
to produce their effect, what is yet to be is thought of as 
being already there and to that we give the name of sprout 
and it is presented as the agent of the action of being born. 
There is further secondary usage when the agent becomes the 
object of the action of attaining one’s form which is the mean¬ 
ing of being born. The agent which thus becomes the object 
through secondary usage is said to be born, that is, it is the 
agent of a process the parts of which are arranged in a sequ¬ 
ence. Explained in this way, the expression ankuro j ayate 
is seen to rest in secondary Being. The agent, the object 
and the action are all the result of upacdra: There is kartr- 
kalpana, karmakalpand and kriyakalpand .] 

It is now pointed out that the expression ‘asti’ which 
expresses the state coming after being born, is also based on 
secondary Being. 

III. 3.47-43 


47. One says about a thing that it exists when it 
'maintains itself’. As the object (of the action of main¬ 
taining) is included in the meaning of the root, the 
latter is not a transitive one. 

[Similarly, the expression ankuro‘sti’ can be explained 
only on the basis of this secondary Being. When one 
says: ahkuro‘sti’ £ the sprout exists’, what is meant is that 
the sprout maintains itself. In this meaning of the root 
‘as’, the agent and the object of the action are the 
same. The agent itself becomes the object (karma) in the 
explanation of the meaning of the root itself. That is why 
the root ‘ as’ is said to be intransitive.] 

.48. Before connection with existence, how can there 
be primary Being ? What does not exist cannot be the 
the agent of 'to exist’. But there is secondary usage as 

[The verb ‘to exist’ denotes existence conceived as a pro¬ 
cess, the parts of which are arranged in a sequence. In this 
state, there cannot be an agent in a finished form. Nor can 
there be the action of existence if there is no connection with 
the agent of that action. Nor does one see any other acces¬ 
sory here. Being itself cannot really become both accessory 
and action. The accessory has to be something that is com¬ 
pleted, accomplished ( sicldha ). What is in the process of 
being ‘ sattci’ (existence) cannot be something that is complet¬ 
ed. What has already assumed existence cannot be a process 
(sadhya). Therefore, what has attained its form (born) is 
conceived mentally as having existence, the parts of which 
are arranged in a sequence. The first part of it is the agent, 
the latter part is a process, the object is included in the action 
based on both. Thus, as in the case of birth, the expression 
'it exists’ is based on an agent which has only secondary Being. 
Thus the expressions asti, ndsti and jdyate have been ex¬ 
plained on the basis of secondary Being. These stand for 
three cut of the six transformations of Being (sad bhavavi - 


kcirah ). The other three are included in them. What is call¬ 
ed growth or increase of parts is included in birth. 
What is called change or the appearance of new pro¬ 
perties is also included in birth. What is called decay 
(apaksiyate) is included in destruction (nasti ). As all actions 
come under the six transformations of Being, they have all 
to be explained on the basis of secondary Being.] 

It is now stated that this secondary Being is all-pervasive. 

49-51. No meaning of a word can go beyond this 
secondary Being which is not opposed to things which 
are opposed to one another, which is resorted to by 
different words in order to present difference and 
opposition, which has no temporal distinction but exists 
in things belonging to different times, which is the 
cause of the use of all words. It has been shown as 
something different from primary Being in the Bhdsya. 

[Things which are opposed to one another have no oppo¬ 
sition in this secondary Being. It is on the basis of this that 
words present difference where there is no difference in rea¬ 
lity. For example, words present the jar and its blueness 
as two separate things, whereas in reality, the two exist as 
one concrete thing. Words like past and future present things 
which are opposed to and different from what words like 
vartaraava denote. This secondary Being presents as sepa¬ 
rate what is really united, but only to unite it again. It ex¬ 
tends to all things, present, past and future. It is the cause 
of the use of all words, because we first conceive things in 
our mind and then give expression to them in words. What 
has not yet been mentally conceived cannot be expressed. 
Even what is actually present cannot be expressed unless it 
has been conceived by the mind. To conceive of things as 
related to one another by the relation of action and accessory 
is also a case of seconadry Being, because it is the work 
of the mind. When we utter a sentence consisting of nouns 
and verbs and expressive of an action which is being done, 

In. 3.52 


the action is not there yet and nothing can, therefore, be an 
accessory in realtion to it. Therefore, the meanings of the 
verbs and nouns in that sentence have an existence only in 
the mind. Thus, all words move in the realm of this secon¬ 
dary Being. What is called non-existence is also something 
which can be conceived by the mind. No word is separated 
from its meaning understood as something moving in this 
secondary Being*. This is accepted by the Bhdsyakdra when 
he says: na sattdm padartho vyabhicarati (M. Bha on 
P. 5.2.94). The past and the future have no outside reality, 
but they have this Being. When we use the word ‘asti ’ in 
regard to a thing, what we are doing is to say that it has 
outside reality in addition to having secondary Being. Even 
external Being becomes capable of being expressed by words 
only when it is grasped by the mind. Past and future Being 
have an existence only in the mind.] 

See Kaiyyata’s Pradlpa on the Bhasya sentence “ na 
sattdm jiadartho vyabhicarati”, on P. 5.2.94. What is stated in 
detail on £ upacdrasatta 9 in this section of Helaraja’s commen¬ 
tary is stated there briefly. 

It is now shown that all words positive or negative are in 
the same position. 

52. Verbal communication relates only to a part 
of an aspect of reality or to the determination by means 
of an external factor or to a reversal of reality or to an 
absence of it. 

[No word expresses reality faithfully. Words like ‘past’ 
and ‘future’ also express things which do not exist and so do 
not express reality faithfully. Thus words expressive of posi¬ 
tive and negative entities are all in the same position. No 
word can express the full reality. All words express only 
a part of it, because everything in this world is only a part 
or an aspect of the Ultimate Reality. In fact, no word can 
fully express even that part. A word like gliata cannot express 
even that part of reality fully. It just expresses a part of 

v-m. 14 



that part, namely, the universal in it. Anything else which that 
thing may possess requires another word to express it. If 
it is red in colour, the word rakta has to be used to express it. 
A thing can be expressed in words only through some such 
property as the universal which exists in it. A thing as such 
cannot be expressed at all. The One Brahman is presented 
by words as many on the basis of different limiting factors or 
they present what is within us as external to us. They are 
responsible for viparitakhyati. According to the Siinyavadins, 
there is no external reality at all. The different forms which 
occur in our consciousness cannot have reality. They are 
mere dreams and it is these unreal forms which words present. 
In other words, they present what does not exist ( abhava ). 
As words can present only one of these four distortions of 
reality, as stated by (1) grammarians (2) samsargadarsana 
(3) Vijnanavadins (4) Siinyavadins, there cannot be an eter¬ 
nal relation between words and the actual reality.] 

53. Just as, through a defect in our senses, objects 
produce cognitions in which something is superimposed, 
as it were, such is the nature of cognitions producd by 

[In this matter of presenting reality in a mutilated form, 
words are like defective senses. With a defective eye, one 
sees two moons instead of one; with jaundice, one sees the 
conch as yellow. Objects produce cognitions in which a form 
different from what they actually have figures. Words also 
produce a cognition in which a non-existing form figures.] 

54. Words are based on cognitions which do not 
reveal the full reality and so present things in another 
form, not determined by their real form. 

[Effects correspond to their causes. The cause of words 
is our cognitions of things (nirupanapratyaya) . Cognitions 
are in the nature of mental constructs (vikalpa) which never 

III. 3.55-56 


perceive things in all their aspects. How they present only 
a part of reality was mentioned before (verse 52). Due to 
the relation of causality, the cognition which arises from a 
word is a vikalpa. Our determinate cognitions are based on 
words and vice versa. Therefore, due to a certain incapacity 
born of avidya, our determinate cognitions are not capable 
of seeing things as they are. They see them in an unreal 
form according to all thinkers and words convey these un¬ 
real forms. All words do this, words which convey positive 
things like ghata and those which convey negative things 
like ‘past’, ‘future’ ( atlta , andgata.) ] 

55. The child and the scholar, when they are both 
at the worldly level, see things and communicate them. 
Therefore, they are in the same position as far as cogni¬ 
tions and words are concerned. 

[It might be said that those whose vision is not blurred 
and who, therefore, see the full reality express it through 
words and we learn words from them. Why should not words 
then express the full reality? The answer is: What is called 
social intercourse means exchange of ideas. Therefore, even 
those whose vision is not blurred are not different from ordi¬ 
nary people at that level. They also observe objects and give 
expression to what they see. In other words, their words 
also express their determinate cognitions and, therefore, 
present only a part of reality.] 

As cognition always relates to limited objects, in what 
its purity consists is now stated. 

58. Purity of knowledge consists in its embracing 
all objects and not having (sense-contact as its) basis. 
When no form of objects figures in it, purity, some say, 
reaches a still higher stage. 

[When sages go beyond the worldly level, there comes 
•<a certain purity in their cognitions. Purity is of two kinds, 



initial and final. The knowledge of the omniscient which 
embraces all objects and which does not arise through sense- 
contact has initial purity. Dependence on the senses is itself 
an impurity. In its finished form, it is free from the appear¬ 
ance of the forms of objects or of any differentiation, it is 
pure Consciousness like the sea without the slightest ripple 
on its surface. It is the supreme Brahman.] 

57. When knowledge exhibits the forms of external 
objects, it is a kind of disturbance, a kind of impurity 
consisting in its being mixed up with them as a result 
of contact. 

[When pure consciousness comes into contact with 
objects and gets mixed up with them, a certain impurity 
results, a certain proneness towards outside objects, just as 
pure water is tainted when it comes into contact with dust. 
When this impurity goes away, the consciousness is restored 
to its original purity.] 

58. Just as knowledge becomes impure through 
being coloured by the object, in the same way, an object 
falls form its real form when it is expressed through 
some limiting factor. 

[The object also is susceptible to impurity. A thing can 
be spoken of only through some characteristic of it, some 
limiting factor such as the universal. This is a limitation of 
the object and so it constitutes a kind of impurity of the 

59. The object, the word and the cognition being 
thus distorted, existence and non-existence thus enter 
into verbal or worldly usage in an identical manner. 

[The impurity of the object consists in its being coloured 
by the universal and the like, that of the word in not being 

III. 3.60-61 


able to express the whole object, but only as coloured by 
some limiting factor or other and that of cognition in being 
coloured by the external object. Thus any positive entity, 
when expressed by words or cognised by the mind, appears 
in the form of some limiting factor. In this respect, it is like 
a negative entity, which cannot be cognised as such. It can 
be cognised only as the negation of something. In other 
words, its nature is determined by something extraneous to 
itself. It can never be cognised in its own real form. Even 
though an error is a defect of the perceiver, the fact remains 
that, in an error, three things, the word, the object and the 
cognition appear in an erroneous form.] 

60. Just as the non-existence of a thing is cognised 
on the basis of its existence, similarly the existence of 
a thing is conceived on the basis of its non-existence. 

[Non-existence, unless related to some positive entity, 
cannot enter into usage. Positive entities also exist only 
during the middle stage, between previous non-existence and 
later destruction. Even eternal things are as good as non¬ 
existent till they are manifested. Being thus mutually de¬ 
pendent, they are similar in their position in worldly usage. 
All this has been said on the assumption that positive and 
negative entities are two different things.] 

61. It is not non-existence that is transformed into 
existence nor existence which becomes non-existence. 
Existence and non-existence are two appearances 
( vikalpita ) and are not different from the one Self. 

[According to dualists, existence comes into being after 
having destroyed non-existence and non-existence comes into 
being after having suppressed existence. This position is 
untenable, because existence and non-existence can merge 
into each other. If they were totally different from each other, 
one cannot be cognised in terms of the other. The fact is 
that the Self, the only Reality, manifests itself now as bhdva 



and now as abhava which are thus two limiting factors of 
one reality. This only reality is in the nature of light and 
always remains so. When, through nescience, it manifests 
itself as the external object, is associated with the present 
time and has causal efficiency, it is said to be a positive entity 
( bhavci ). When it is associated with past and future time, 
it is in the form of a residual trace only and then it is said 
to be a negative entity ( abhava ). But it is not absolute non- 
being because it is remembered if it is associated with past time 
and imagined if it is associated with future time. In this condi¬ 
tion, it is not perceptible to the senses. There is no non-exis¬ 
tence which has not even got the form of residual trace 
(samskdra) and is devoid of all powers. Things are either 
present and can be perceived by the senses or they are past 
or yet to come and are, therefore, remembered or imagined. 
That is why things are said to be traiyyaclhvikdh, that is, 
belonging to three paths, the present, past and future. Thus 
absolute non-existence does not exist.] 

The untenability of the relation of causality is now stated. 

62. Non-existence being intangible, cause cannot 
bring it about. Nor can cause do anything to what is 
already tangible. 

[It might be said: why look upon existence and non¬ 
existence as mere mental constructs? Why not look upon 
them as two different effects, due to different causes? The 
reason is that the relation of causality itself in untenable. 
Causes are supposed to act in such a manner that effects come 
into being. How can causes act keeping in view something 
which does not exist yet? It is true that it exists in the form 
of the universal, but the universal, being eternal, cannot be an 
effect. Nor would it do to say that the universal and the 
particular are identical and as the particular is not eternal, 
it can be an effect Because how can the causes work keep¬ 
ing the particular in view? Moreover, if the causes work for 
the coming into being of the non-existent effect, why should 
the latter come from one particular cause rather than another. 

HI. 3.63-64 


Why should oil, for instance, come from sesame-seeds and 
not from sand, considering that it was non-existent in regard 
to both? Nor can one argue that the power of causes to pro¬ 
duce effects is limited. We see that when something is there, 
something else is produced and not otherwise. Well, it is 
only a matter of our understanding such a connection be¬ 
tween two things. There is really speaking no such connec¬ 
tion. Satkdryavdda is also not tenable. If the effect is already 
there, the cause cannot be said to produce it. The idea that the 
already existing effect is merely manifested by the cause is 
also not tenable. The manifested effect is either different 
from the cause or it is not. If it is different, it either existed 
or did not exist before manifestation. If it did not, then 
there can be birth of what did not exist and that is not possi¬ 
ble. And so on. The net result of this argument is that 
causality as understood by others is not possible and that 
it is based on the notion of existence and non-existence as 
two distinct things.] 

63. Therefore, everythng is either existence or non¬ 
existence. There is no other state which proceeds from 
the one Reality. 

[The fact is that they are not two distinct things. They 
are only two unreal appearances of one ultimate Reality 
( vivartd ). due to nescience. As this does not appear as it 
is, the whole of the phenomenal world is a kind of non-exis¬ 
tence, a'paramdvthci.'] 

64. Those who believe in a positive reality do not 
accept non-existence and those who believe in non¬ 
existence only do not accept any positive entity. 

[The sages hold the view that birth and death are 
nothing more than the manifestation and the hiding of the 
real. What is called non-existence is not something distinct 
from existence. It is nothing more than the previous and 
later conditions of existence. The previous non-existence of 


a jar is clay and its later non-existence is potsherds (fea- 
pala). The Sunyavadins consider non-existence alone to be 
the truth. But even they have to admit existence at the 
worldly level. Thus through ‘samvrti’ or hiding of reality 
at the worldly level, everything can be explained.] 

How the distinction between existence and non-existence 
arises, is now explained. 

65. While all this visible world is naturally one, 
there is an extraordinary order in this multiplicity. 

[All this variety is really pervaded by a certain unity, 
namely, Brahman. But due to nescience, this unity appears 
as multiplicity. It is this very unity which appears as exis¬ 
tence and non-existence; but, of course, this distinction is 

66. Just as four states are postulated in what is 
intangible (nirupdkhya), in the same way is this two¬ 
fold division into existence and non-existence postulated. 

[Ix non-existence which is intangible can have such 
artificial divisions as previous non-existence, non-existence 
after destruction, absolute non-existence and mutual non- 
existence, the ultimate Reality also can have artificial 
distinctions. What is called existence is the fact of being 
cognisable to the external senses and fulfilling some purpose 
or other. What is called non-existence is the fact of being 
imperceptible and not fulfilling some purpose.] 

It is now stated that neither existence nor non-existence 
has separate Being. 

67. Non-existence can rationally be neither opposed 
nor non-opposed (to existence) can be neither existent 
nor non-existent, neither have sequence nor not have it. 

[As non-existence in intangible, there cannot be opposi¬ 
tion between it and existence. Non-existence neither helps 

III. 3.G8-71 


nor hinders existence. One cannot predicate existence about 
it, because what exists cannot be non-existent and what has 
already attained its form cannot get anything further. Nor 
can it have a separate form because, in that way, there would 
be no difference between it and existence. Sequence is an 
attribute of existent entities and, therefore, non-existence 
cannot have it. Nor can there be absence of sequence, that 
is, simultaneity, because that depends upon something ex¬ 

68. Existence which is opposed or otherwise, 
existent or non-existent, sequenceful or sequenceless, 
does not, therefore, exist. 

69. There cannot be any division within non¬ 
existence on the basis of the three divisions of time 
and, if that is not possible, there cannot be triple time 
for existence either. 

70. Abandoning one’s own essence due to something 
external is not possible, nor is it possible to maintain 
that one’s own essence depends upon oneself or some¬ 
thing external. 

71. There is contradiction in identity (between 
existence and non-existence) and if they are different, 
they cannot render service (to each other). If both 
are abandoned, all usage would cease. 

[Verses 67-71 are intended to show on the basis of logic, 
that really speaking it is not possible to show the validity of 
existence or non-existence. They are no more than appearance 
and disappearance or hiding of something unreal. The ulti¬ 
mate Reality is eternal and is not affected by its appearance 
and disappearance as unreal forms of something which is 
eternal and quite free from these unreal forms.] 

The truth is now stated. 

V-III. 15 


72. Those who know the final portion of the Vedas 
have declared that entity alone to be real which is 
differentiated into the see-er, the seen and the seeing. 

If the text is: ‘va avikalpitam’, the translation should be: 
in which there is no real differentiation into’. 

[Knowers of Brahman have declared that the world 
consisting of the perceiver and the perceived, brought about 
through differentiation is not real. What is real is beyond 
all differentiation, beyond the range of words and mind, with 
no beginning and no end. The final portion of the Vedas, 
consisting of Rk, Yajus, and Savian, is the TJpanisad where 
the essential truth is propounded. Those who know this 
portion have declared that monism is the truth and not dualism 
suggested by the performance of actions. Differentiation is 
due to avidya which is neither identical with nor different 
from the ultimate reality.] 

It is now stated that words move about in the world of 
unreal differentiation. 

73. Inasmuch as words express the universal or the 
particular as differentiated, they move about in the 
world of unreal differences. 

[Words express the unreal which appears in our cogni¬ 
tion and do not touch the undiffei’entiated reality but move 
about in the world of differentiation. Words express even 
universals as distinct from other universals. That is, even 
when they express universals which stand for unity, they are 
still associated with differentiation. Similarly, a visesa (par¬ 
ticularity) is dfferent from other vsesas, as expressed by 

The author now says somethng about non-existence as 
understood by the Vaisesikas. 

74. When non-existence is brought about, existence 
cannot be destroyed nor can existence be brought when 

III. 3.75-77 


non-existence is destroyed (if non-existence is also a 
separate category). 

[So far, all reasoning has been going on on the basis of the 
view that non-existence is something intangible. But, for the 
Vaisesikas , it is a separate category. That means that it is 
something tangible. If that is so and it is different from 
‘bliava’, how can bliava be destroyed when abhdvci is pro¬ 
duced and how can bliava be produced when abhdva is 

75. The existence of sdbaleya does not prevent that 
of bdhuleya, nor does the non-existence of sdbaleya 
bring about the existence of bdhuleya . 

[Just as the existence of sdbaleya does not prevent that 
of bdhuleya and vice-versa, in the same way, if abhdva is also 
a separate category, its existence or non-existence would not 
bring about the non-existence or existence of bhdva.~\ 

76. Such considerations would arise if non¬ 
existence were a positive entity and in regard to its own 
non-existence, the whole question would arise again. 

77. Therefore, being a non-entity, it is beyond the 
scope of verbal communication and variation within it 
does not deserve to be explained. 

[If we proceed on the Buddhist assumption that what is 
called non-existence is only the disappearance of a positive 
entity, and, therefore, something intangible, all these con¬ 
siderations would not arise at all. The objection that when 
non-existence is destroyed, why should a positive entity come 
to be, would not be raised. If the object comes into existence 
when its previous non-existence is removed, it is because of 
its nature to come to be. Therefore, non-existence is some¬ 
thing intangible and does not deserve to be classed into a 



separate category. It is intangible and it cannot be thought of 
in the same way as bhavci .] 

The untenability of causality, already stated, is now 
repeated and vivartavcida, the final doctrine, is expounded. 

78. There is no such thing as the cause working 
towards something which has not even a word to express 
it. Even when the cause is present, there is the absence 
of the effect (before it is born). 

[As the effect does not exist before it comes into existence, 
it has no word to express it. The cause, therefore, cannot 
work towards it. A cause operates in relation to an effect 
which is fit to be born and which is something positive. So 
unless the effect exists already, the cause cannot work towards 
it. Causes do not work towards something which is nothing. 
That which does not exist cannot be an aim and if it already 
exists, there is no need for a cause. If, by what is called the 
cause working towards the effect, it is the agent’s plan which 
is meant, that is also not possible. An insentient cause can 
work still less towards the effect. So what is called causality 
is nothing more than something appearing to come into 
existence only when something else is present.] 

79. ‘What is its previous condition’ is a question 
which relates to something which exists. Both the ideas 
of ‘previous’ and condition are inapplicable to some¬ 
thing which does not exist. 

[No one can say that because, before its production, there 
is its previous condition, therefore, it is not right to say that 
there is no word to express it. Because, one can legitimately 
ask ‘what is this previous condition’ only in regard to some¬ 
thing which exists. Neither the idea of previous nor that of 
condition applies to something which is nothing. Something 
which is devoid of form or shape cannot be said to be previous’ 
either spatially or temporally. The idea of condition pre- 

III. 3.80-82 


supposes one who is in that condition and no such thing is 
visible. Thus, both according to asatkdryavdda and satkdrya - 
vdda, there cannot be a previous condition of the effect.] 

80. After destruction, one cannot say about a thing 
that it exists or does not exist, because of the absence 
of any ground for doing so. It is beyond the range of 

[After destruction, one cannot say of a thing whether it 
exists or not. The Sankhyas affirm that it does exist, but that 
is not right. The fact of being cognised is the basis of such 
statements and that cannot happen with something which has 
been destroyed. Therefore, before it is produced and after it 
is destroyed, one cannot speak about a thing at all.] 

It is now siated that it is only in the middle stage that 
it is visible. 

81. It is indeed a wonderful activity by virtue of 
which the indivisible and sequenceless essence of pre¬ 
viously non-existent things manifests itself. 

[Causality has already been shown to be unsound. So 
the effect, previously non-existent, suddenly appears under 
certain conditions, that is, when something else is present. It 
only appears to be an effect, but it is not so. People, blinded 
by nescience, imagine it in many ways. It is only a vivarta. 
When the one, without ceasing to be one, assumes different 
unreal forms, there is vivarta.] 

Helaraja here quotes the definition of vivarta found in 
the vrtti on Vak. I. 1. 

82. All worldly usage is carried out with objects 
created by mental construction as with primary objects. 

[After having seen the object during the middle stage, 
one imagines a previous and a posterior stage for it and thus 



it becomes endowed with sequence and then one postulates a 
relation of causality and talks about blidva and abhdva. This 
is worldly usage which is the result of nescience.] 

83. Those who believe in the Eternal (Brahman) 
look upon this as the power of the only Reality. They 
declare sequence to be the same as the objects and not 
as something apart from them. 

[In reality, the plurality which is visible is not the truth. 
Only monism is the truth. It is seen as plurality by those 
whose eyes are dimmed by nescience. The one sequenceless 
Brahman appears as many and as possessing sequence because 
of the power called svdtantrya. These two powers, that by 
which the one appears as many and the sequenceless appears 
as having sequence are not really different from Brahman. 
Nothing besides Brahman can shine and that which shines 
cannot be different from Brahman. Sequence appears as 
conditioned by objects which are identical with Brahman. So 
sequence is really not different from what has sequence.] 

84. In reality, simultaneity is not different from 
sequence just as non-existence is not different from 

[The opposite of sequence is simultaneity and that is also 
a mere appearance. The One Reality cannot be cognised as 
such by ordinary people. It manifests itself as having spatial 
and temporal sequence. When two or more things appear 
together, they seem to have simultaneity. But squence and 
simultaneity are not different from the things themselves. It 
has been said before that abhdva is not different from bhdva 
and that it is not anything positive. Simultaneity is only 
absence of sequence. Association in time with something else 
is not different from that which so associates.] 

85. People of the world speak about time within 
time. But one cannot bring about a distinction by 
merely speaking about it. 

lit 3.86-88 


[By speaking about time which exists to-day and time 
which will exist tomorrow, some people make a distinction 
within time. They speak of time as ddhdra or as adlieya. But 
Time is One. Similarly, sequence and simultaneity do not 
differ from each other nor from the things which are 
sequential or simultaneous.] 

86. Even if non-existence is imagined to be a 
substratum, nothing can really exist in it. Nobody can 
be prevented from imagining totally non-existent things. 

[When we say salvor abhdve sukham the use of the loca¬ 
tive case seems to imply that non-existence can be a substra¬ 
tum. .But that is not right. It is a mere fiction. There is 
nothing which one cannot imagine.] 

87. Therefore, the one eternal Reality, consisting of 
existence and non-existence, shines through its different 
powers, in many forms when conveyed by words. 

[The substance of this section can be stated as follows — 
Ihe ultimate Reality is beyond all differentiation. It is 
endowed with all powers. Words express this Reality, accord¬ 
ing to occasion, either as a positive entity or as a negation 
and as having this or that limiting factor. Those words which 
express negative things are similar to words expressive of 
positive things as far as their relation to their meanings are 
concerned. This ultimate Reality is one, because there is no 
differentiation within it. It is eternal, because there is no 
such thing as abhdva, non-existence.] 

88. Verbal communication in the world is done 
with meanings of words fashioned by the mind and in 
the science of grammar, it is the meanings adopted in 
the world on the basis of which the work (of explaining 
the forms) is done. 



[Therefore, in everyday life, it is the norm for words to 
express the limited reality and it has been so stated in this 
sastra. Therefore, all the different views of the tarkxkas are 
out of place. Ordinary people do not follow the conclusions 
which may have been reached by different thinkers. In the 
matter of using words in accordance with our cognitions, both 
the scholar and the child are equal. On that basis, they pro¬ 
ceed to exchange views. Ordinary people understand things 
in a superficial manner and use words accordingly. This sastra 
is an attempt to explain words as used by ordinary people. 
In this attempt, one should take the help of the ordinary 
man’s view of the world. From the indivisible sentence¬ 
meaning, by a process of abstraction, one derives the root 
expressive of action, the nouns expressive of things, the suffix 
which conveys its own meaning for the sake of explaining the 
forms. 'The notions of action, substance, quality and so on 
used by the science of grammar, are worldly notions. In 
worldly usages, all words whether expressive of positive or 
negative things, are in the same position. Therefore, the 
meanings of words have continuity. Thus the relation between 
word and meaning is eternal.] 

III. 4,1-2 


Section 4 


1-2. Some meanings of words which exist in the 
sentence looked upon as interconnection ( samsargaru - 
pat sambhutah) which are isolated by a process of 
abstraction (scimvidrupad apoddhrtdh ) which are 
separated from the meaning of a sentence like the 
meanings of the stem and the suffix (from that of a 
word), which are the basis of the correctness of words 
and which are inferred from indications in the Science 
of Grammar, will now be explained according to 

[The previous section ended with the remark that, in the 
Science of Grammar, meanings of words, agreeing with 
worldly usage, are isolated for the purpose of explaining 
the formation of words. The sentence is indivisible and so 
is its meaning. For the purpose of conveniently explaining 
the forms of the language both are artificially divided. These 
divisions become the means of the derivation of the forms 
of the language. For the purpose of this division, the mean¬ 
ing of the sentence is looked upon as connection (samsarga ). 
In that way, the meanings of the individual words whose 
connection the sentence-meaning is, can be separated. They 
exist only if the sentence-meaning is looked upon as a com¬ 
bination. If they do not exist at all or if they are like bits 
of iron, they cannot combine and there cannot be any ques¬ 
tion of analysis. Of course, the sentence and its meaning 
are indivisible. So there are no word-meanings. There 
cannot be any question of their previous separate existence. 
The hearer does not understand the meaning of the whole 
sentence all at once. He understands it little by little, part 
by part and then joins the parts together. The taste of 
‘sharbat’ ( pdnaka ) is really an indivisible whole, but those 
Y-IIL 16 



who drink it can, if they make an effort, taste each ingra- 
dient separately and assess the part played by it in making 
up the taste of the whole. As the indivisible sentence—mean¬ 
ing cannot be understood in a flash all at once, the unreal 
woi d-meanings are abstracted in the middle as mere means to 
an end. Once the sentence-meaning is understood, they dis- 
appear. This artificial division is done with a purpose. It 
is to facilitate the teaching of forms, so that each form may 
be made in its own meaning. These meanings are not 
taught in the sdstra, they are not ‘vidheya’. They are natu¬ 
ral and they are made use of to teach forms. This is the 
meaning of the M. Bha. statement: svdbhdvikam ity alia. 
Kuta etat. avthanadesanat—* It is declared that it is natural. 
Why so? Because meaning has not been taught.” (M.Bha. I, 
p. 363, 1.8). The mention of meaning in a grammatical siitra 
is only by the way. It is only a means to an end. The 
artificial division of a sentence is like the division of a word 
like a noun or a verb into parts and assigning a meaning to 
each, though a word is really indivisible. These artificial 
meanings are indications of the correctness of the word. That 
is why they are analysed. They are of two kinds: (1) those 
which are known in the world, (2) those which are known 
in particular sdstras. These latter are defined by gramma¬ 
rians in their own way. These definitions are inferred from 
Panini’s practice. For example, from the siitra which tea¬ 
ches ‘ekasesa’ (P. 1.2.64) we gather that the individual or 
the particular ( dravya ) can be the meaning of a word and 
that it means something which is to be differentiated 
(bhedya). From the vdrttika — “yasya gunasya hi bluwdd 
dravye sabdanivesali etc.” (Va. 5 on P. 5.1.119), we under¬ 
stand that guna is something which rests on something else. 
Even the universal (jdti) can be said to be guna when it 
is expressed by an abstract suffix (bhdvapratyaya). Words 
like sukla in suklah patali’ denote an object having white 
as its quality. That what is called time is essentially action 
is known from the fact that Panini has used such words as 
‘bhuta’, ‘vartamdne’, ‘bhavisyati’ as expressive of the limit¬ 
ing factor of the meaning of the root which is action. That 

III. 4.3 


number is something which enables us to count difference 
is made clear in the sutra-jatyakhyam ekasmin’ etc. 
(P. 1.2.57). The Vaisesika conception of number is that it 
is a quality which inheres in a substance. We gather that 
samstydna, prasava and sthiti constitute the nature of the 
genders from the expressions ‘ striyam\ *pumsi and ‘napum- 
saklie’ which are found in connection with the explanation 
of the forms of words. Gender cannot mean sex because 
such things as a ‘khatva’ cannot have sex in the ordinary 
sense of the term. That it is power or capacity ( sakti ) 
which is the real accessory ( sadhana ) and not substance can 
be gathered from these indications: (1) from the fact that 
an accessory can sometimes be sesa (2) from the fact that 
the object can sometimes become agent — changes which 
substance, being uniform, cannot undergo — (3) from the fact 
that an ‘avyayibhava’ compound has been taught in the sense 
of a case-ending. The concepts of ‘pui'usa’ and ‘upagraha* 
have been taken from previous grammarians. Space and 
action are well-known in the world. Action has been de¬ 
fined as a process the parts of which are arranged in a tem¬ 
poral sequence. One and the same word can convey many 
of the things mentioned above, but one of them as the main 
thing and the others as subordinate to it. Thus, the verb 
denotes action, time, accessory, number, person and aspect. 
The noun denotes substance, gender, number, accessories, 
action and time. All these things have been explained ac¬ 
cording to the tradition of the grammarians.] 

As quality etc. depend upon substance, the latter is first 

3. That in reference to which a pronoun can be 
used is substance, presented as something to be 

[Pronouns can do one of two things. Some merely de¬ 
note things in general, like K sarva \ Others denote particular 
things like ‘ anyatara It is the former which are used to 
refer to substance. In fact, that is just the characteristic 


of substance, namely, that it can be referred to by a pro¬ 
noun (sarvandmapratyavamarsayogyatvam ). Pronouns re¬ 
fer to things in general either as present or as past. When 
the element of present or past is discarded, what remains is 
just the thing in general and that is substance. It is present¬ 
ed as something to be qualified by such limiting factors as 
the universal. What is meant by ‘presented’ is that what is 
being defined is not external reality, but reality as presented 
by words. Thus, even a universal, when presented as some¬ 
thing to be qualified, becomes substance. This conception 
of substance is quite different from that of other schools. In 
this way, anything can be presented by words as substance, 
as something to be differentiated. This is true of the other 
conception also, namely, that of Vajapyayana who held that 
all words denote the universal i.e. something which is the 
cause of the same cognition arising and the same word being 
applied to something. Such a universal exists in action 
also. That is why we cognise every step in the pro¬ 
cess of cooking as cooking. If only the last stage in 
the process, that is, the softening of the substance 
cooked, were the meaning of the root ‘to cook’ (pac ), 
then the other stages would have to be understood 
through inference or implication and would not be the 
actions of the cook. If the activity of the agent is not ex¬ 
pressed by the root, but is understood by inference or impli¬ 
cation, then there would be no connection between the mean¬ 
ing of the suffix, namely, the agent and his activity which 
would not be expressed by the root but only understood by 
implication. If the root ‘pac’ denotes only the softening of 
the material (viklitti ), then the material can only be kcirta 
or kcmnakarta , but never the object. In other words, we 
would never be able to say — devadattcih odanam pacati. 
Secondly the softening is only the last stage or moment. So 
it cannot be the action at all. Action is a process of which 
the parts are arranged in a sequence. The action of cooking 
is in the same position. Such moments as putting the vessel 
on the fire must be part of that process, because at these 
jnoments also ..we get-the idea of cooking. So there is such 

in. 4.3 


a thing as the universal of cooking inhering in every moment 
of the process. This is true of all actions. Even though 
every verb denotes the universal aspect of action, the acces¬ 
sories become connected with the individual aspect which 
is something to be accomplished ( sddhya ) and only sddhya 
can be connected with ‘ sadhana ' (the means). 

Just as it has been shown that action is a universal, it 
can also be shown that it is a substance. Even according 
to the view that the verb denotes action primarily, it must 
denote substance, that is, the accessory secondarily. Other¬ 
wise, in regard to what would action be the primary mean¬ 
ing? Would it not be simpler to say that the verb denotes 
primarily the accessory in activity? Even according to the 
other view, the accessory comes in secondarily. Otherwise, 
the expression ‘devadattah pacali where the word expressive 
of action and the word expressive of accessory are in appo¬ 
sition to each other, would be inexplicable. Action, the 
meaning of the verb, can be referred to by a pronoun, charac¬ 
teristic of substance. In the expression c sadhu joacati’ we 
see that the meaning of the verb is modified or differentiated 
by the word ‘sadhu’. Similarly, the verb ‘dsyate’ can be 
qualified by the word ‘sukham’ and so its meaning becomes 
‘bhedya’, a characteristic of substance, as stated in the verse 
under consideration. 

The Mimamsakas argue that the verb expresses ‘bring¬ 
ing about’ ( bhdvand ) and that substance is understood by 
implication only and is not the expressed meaning of the 
verb. This is wrong. It is the root which expresses blid- 
vand. The suffix denotes the accessory. Thus the former 
expresses the sddhya and the latter the ‘ sadhana’ and that 
is how the two are correlated. If ‘ sadhana’ is understood 
only by implication, there would be no relation between 
-the two].. 

_ ^ Now the definition of quality (gfitna), based on indica- 
. tions found in the sastra, is being given. 



Section 5 


1. Whatever rests on something else ( samsargi) 
differentiates it ( bhedaka) and is understood in that 
function ( savydpara ), is, being dependent, called 
quality’ in the sastra. 

[Quality is said to rest on its substratum, because the 
forms of the two are mixed up, as it were. That is why it 
distinguishes its substratum from other substrata. The 
force of the repetition of the relative pronoun (yad yad) is 
that anything, even a universal, can become a quality, if it 
is understood as something which distinguishes the substra¬ 
tum. This is the implication of the M.Bhd. on Vci 21 on P. 2.1.1. 
This definition of guna follows worldly notions. Va 5 on 
P. 5.1.119 also stands for the worldly notion of guna. That 
varttika says something about the meaning of the suffixes 
‘ tva’ and ‘taV when added to stems. If the stem stands for 
something to be distinguished (visesya), these suffixes ex¬ 
press the distinguishing quality (visesana ). When a stem 
like ‘go’ stands for the individual characterised by class or 
universal, the suffix added to it stands for the class or uni¬ 
versal. When the stem stands for the class only, then the 
suffix stands for the form of the word itself. That the form 
of the word also is the expressed sense of the word has al¬ 
ready been explained. It can also be looked upon as a qua¬ 
lity which qualifies the meaning of the stem. It is super¬ 
imposed on it, it is identified with it. When a stem like 
‘sukla’ stands for the individual qualified by the quality 
‘white’ the suffix ‘tva’ added to it stands for the quality. If 
the stem stands for the quality the suffix denotes the uni¬ 
versal inhering in it. If it stands for the universal, the suffix 
stands for the form of the word as before. Even when the 
suffix ‘tva’ is applied to a proper name like ‘Dittha’, it de¬ 
notes the universal, that element which persists in the indi¬ 
vidual named, through all the changes which he undergoes. 

III. 5.1 


In words like rdjapurusatva, pdcakatva, aupagavatva, the suf¬ 
fix denotes relation. The word rajapurusa’ denotes a person 
qualified by a relation with the king and the suffix denotes 
that relation. In ‘ hastitva 9 and kumbhakaratva’ the suffix 
denotes classes, relation being the basis of the formation of 
the stem itself. From all this, it is clear that in P. 5.1.119, 
the word l bhdva’ stands for something which qualifies some¬ 
thing else, due to which a thing appears as what it is (b7ia- 
vaty asmdt tenakdrena clravyam iti ). This conception of 
quality ( guna ) has been adopted by the Science of Gram¬ 
mar. Sometimes, however, rules have been framed on the 
basis of other notions of quality. For example, P. 4.1.44. 
.Mere the word ‘gunavacana’ means a word expressive of sub¬ 
stance to which an attribute is subordinate and the feminine 
suffix comes after such a word. The word guna’ in this stitra 
cannot mean what is defined in the present verse because 
then the feminine suffix would have to be added even to a 
word expressive of the universal. It stands for the Vaisesika 
notion of guna, mentioned in the verse “sattve nivisate” etc. 
given in the M.Bhd. under P. 4.1.44. This definition would 
not include the universal. The other verse given in M.Bhd. 
under P. 4.1.44, namely, ‘upaity anyaj jahdty anyad ’ etc. also 
gives the Vaisesika definition of guna. According to some, 
these two verses are not meant to define ‘guna’ but to say 
what kind of word ‘gunavacana’ is. M.Bhd. on P. 1.4.1. denies 
the name 'gunavacana' to a compound word, a word ending 
in a primary or secondary suffix, a pronoun, an indeclinable, 
a proper name and a word expressive of universal. It is a 
word which denotes a thing qualified by an attribute. After 
such a word, the suffix ‘syan’ can come according to P. 5.1.124. 
Thus, this siitra is also based on the Vaisesika conception of 
guna, though not in a consistent manner. 

In P. 5.2.94 and P. 8.1.12, a word which denotes a thing 
to which a quality is subordinate, is gunavacana. For exam¬ 
ple, the word ‘sukla’ which means not ‘white’ but ‘that which 
is white’. Or the word ‘patu’ which does not mean cleverness, 
but one who is clever. A word which denotes quality only is 
not ‘ gunavacana’. For example, ‘kdrsnyam’. In P. 5.2.47 the 



word guna stands for an equal part. The meaning of the 
word ‘guna’ in P. 2.2.11 has to be clarified. The vdrttikas 
tatsthais ca gunaili and c na tu tadvisesanaih’, given under 
P. 2.2.8 must be taken together with P. 2.2.11. In the last 
siitra, the compounding of a word ending in the genitive case 
affix with a word expressive of guna is prohibited. There 
the word guna is used in the Vaisesika sense. So we cannot 
have a compound word for kdkasya karsnyam’. This prohi¬ 
bition is set aside in some cases. Words express guna in two 
ways. Sometimes they present it as quite separate from 
the thing in which it exists as in candanasya gandhah \ We 
cannot say ‘candancim gandluth’ because the word ‘gandhah’ 
denotes smell itself and not smell as existing in sandal wood. 
When we say ‘suklah patah’ the word ‘sukla’ denotes ‘white’ 
as an attribute of cloth. It means something that is white 
and not whiteness itself. ‘Gandhah’ is what is called ‘tatstha 
guna’ that is, guna which is presented as something separate 
from its substratum. A word ending in the sixth case affix 
can be compounded with such a word. That is why we can 
say: candanagandhah . A word which presents a quality 
sometimes as separate and sometimes as one with its sub¬ 
stratum cannot be compounded. We can say ‘patasya suklah’ 
and ‘suklah patah \ In the former expression ‘white’ is pre¬ 
sented as something separate; in the latter, it is presented 
as existing in the cloth. Therefore, we cannot have the com¬ 
pound word patasukla. So the word ‘guna’ in P. 2.2.11 means, 
not a word which means a thing to which a quality is 
subordinate, but a word which presents a quality as some¬ 
thing separate and independent, though elsewhere it might 
be dependent. In Va. 1. on P. 2.2.7 and in P. 5.3.58, the 
word guna has the Vaisesika sense. In P. 6.2.155, it has the 
meaning given to it in the present verse. From all this it 
follows that the meaning of the word guna varies in the 
sdstra and that the meaning given to it in the present verse 
is the special grammarian’s meaning.] 

Here certain doubts arise. Why is it that only quality 
admits of degree and not the thing itself? How is it that 
when degree is expressed in regard to quality, degree in 

III. 5. 2-5 


regard to the thing is also understood? How can one say 
that everywhere some quality is present which brings about 
difference of degree in the thing, considering that qualities 
cannot have qualities, universals cannot have universals and 
actions can have no qualities? How can excellence in one 
thing bring about excellence in something else? An attempt is 
made in the following verses to remove some of these doubts. 

2. Whatever distinguishing quality is adopted to 
determine a thing which is otherwise indeterminable, 
any excellence caused by it is expressed (by the 

[A thing cannot be determined except with the help of 
its quality. If a thing cannot even be determined in its own 
form, there cannot be any degree in it except through some 
quality or other. Through it, a thing can be distinguished 
from other things. Only that quality can distinguish which 
is actually mentioned, and not those which are understood. 
There are too many of the latter and one would not know 
which one to take. Degree comes as a kind of excellence of 
the meaning of the stem and only that which is expressed can 
be the meaning of the stem and degree in it is expressed by 
the suffix. In this verse, the word guna stands for the notion 
defined in the previous verse. Even a stick can be a quality, 
because it can distinguish one with it from one who does 
not have it.] 

It is now stated that just as difference in degree in a 
thing cannot exist except through a quality, in the same way, 
when a word presents an attribute as apart from the thing 
in which it exists, as something independent, difference in 
degree in it can be expressed only through another quality. 

3. Without a distinguishing quality, that which is 
(presented as) the main thing cannot admit of degree 
nor come within the range of words. 

V-m. 17 


[That quality through which it comes within the range 
of words is the very one through which difference of degree 
is also expressed. The universal becomes the cause of some¬ 
thing being expressed by words, but not of difference of 
degree in it, because it exists everywhere in the same degree. 
It does not admit of difference in degree. It exists from the 
very time when a thing comes into existence. So some other 
quality co-existing in a thing with the universal, becomes 
the cause of difference of degree in it. Similarly, when a 
word presents a quality as the main thing, difference in 
degree in it can be expressed only through another quality, 
as in ‘suklataram rupam 9 where it is some such quality as 
‘brightness’ ( bhdsuratva ), or purity ( vaimalya ) through 
which degree is expressed. If brightness is presented as 
something independent, that is, as a thing, it would require 
another quality to express degree in it, as in bhasuratara . 
The idea has already been set forth in Vak. 1.64.] 

It is now declared that ‘brightness’ is actually conveyed 
by the word ‘sukla’. 

4. All the attributes present in the (main) thing 
do not serve to distinguish it. They are conveyed by 
special words expressive of some distinguishing mark. 

[It cannot be said that ‘brightness’ through which degree 
is expressed is not directly conveyed by the word * sukla ’. 
The fact is that there is no one word which can express all 
the distinguishing qualities which exist in a thing. Each word 
expresses only one of them. Difference of degree can be 
expressed only through an attribute actually expressed by 
the word. ‘Brightness’ is actually expressed by the word 
‘sukla’. This word cannot be applied to any object which 
is not bright. If we say ‘suklatara’, it is because there is a 
greater degree of brightness. Therefore, one must assume 
that ‘brightness’ is actually conveyed by the word ‘sukla’.] 

The very attribute which brings an object within the 
range of words can also be the cause of the expression of 

III. 5. 5-7 


5. If a special attribute is useful only for making 
a thing namable, then difference of degree would re¬ 
main without a cause to make it known, as there would 
be no restriction (as to which other attribute should 
do it). 

[One cannot argue that, even though a thing may come 
within the range of words through some attribute, the latter 
cannot serve to express degree also. If it serves only to 
bring the object within the range of words, then we would 
see difference of degree being expressed through some other 
attribute such as weight. But that is not what happens. We 
understand difference of degree through the first attribute 
itself as in ' suklatarci\ Therefore, we must conclude that the 
very attribute which gives something a name serves also* to 
express degree in it.] 

It is now explained how expressions like nikrstatarah 
;prakrstatarah 5 are possible. 

6. Everything can differ in degree from everything 
else (of the same kind) either on the basis of an attri¬ 
bute of inferiority or of superiority. 

[When there is excellence of that which is the cause of 
something coming within the range of words, there is excel¬ 
lence of the thing itself. When the cause is of an inferior 
nature, the degree relates to inferiority and when it is of a 
suerior nature, it relates to sueriority. Thus within the same 
kind, there can be difference of degree.] 

Expressions like ‘gotarc i’ are now being explained. 

7. If the object does not require a basis ( nimitta) 
for the expression of degree, its mention, when degree 
is to be expressed, would be useless. 

[In the expression gotara, difference in degree cannot be 
due to the universal and there is no word expressive of any 



quality. Difference in degree is, however, understood. Why 
then insist that in i suklatara 3 it is understood through a 
quality? The fact is that it cannot be understood from the 
object itself. And if the attribute also is not the basis of 
the cognition of degree, there would be no point in mention¬ 
ing it. But it is actually mentioned in suklatara . The reason 
is that the attribute is the basis of the difference in degree. 
In suklatara the attribute is brightness. In c gotara y , it would 
be some such attribute as superiority in carrying loads or in 
giving milk.] 

8. Therefore, quality, performing its function 
( savyapara ) and based on its own difference in degree, 
involves the object ( dravya ) also in such difference 
and distinguishes it from others. 

[The object, in itself, does not admit of any degree. 
Therefore, quality, by expressing quality within itself, serves 
to distinguish its substratum from others.] 

It is now stated how difference of degree in one thing 
results in difference of degree in another. 

9. Just as the formless object is expressed through 
the form of another (its quality), in the same way, 
being itself devoid of difference in degree, it acquires 
it through such a difference in its quality. 

[An object, in isolation from its attributes, cannot come 
within the range of speech. When qualified by an attribute 
it can be expressed in words. It is also difficult to distinguish 
a thing from its attribute. Therefore, excellence of the attri¬ 
bute results in the excellence of the object. It is not confined 
to its own excellence, isolated from the object, nor to that 
of the object, isolated from its attribute.] 

III. 6. 1-3 


Section 6 


1. The words Direction (dik), Means ( sadhanam) 
Action ( kriya ) and Time (Kola), (if taken) as expres¬ 
sive of things, would not refer to their nature as poweis 
of things. 

[Words like Direction (Dik) convey a meaning which is 
an attribute of things having an independent character. They 
do not express independent things. The four words Direction, 
Means, Action and Time stand for power, capacity and not 
for independent entities. They are dependent on something 
else. The Vaisesikas look upon Dik as an independent entity, 
a dravya. That is not right. It is something to be inferred 
from its effect, from the service which it renders to positive 
entities. Words which express an accomplished thing cannot 
convey it directly. Its definition must be made on the basis 
of indications in the sdstra as in P. 5.3.27.] 

The definition of Direction (Dik), based on inference, 
is as follows: 

2-3. Direction is that power which is the cause 
of the opposition of the limit ( avadhi) and the limited 
( pratipddya) of the cognition of straight without the 
help of anything else, of the manifestation of the minor 
universals of action and which is divided through its 


[We make statements like this: ‘This is to the east of 
that’, ‘this is to the west of that’. Here, ‘that’ is the start¬ 
ing point and ‘this’ is what is determined by it One has 
to exolain such expressions. It is the notion of Direction 
(dik)'which explains them. Universals etc. which exist in 
objects cannot explain them. As we have to infer its nature 
from its effects and as, in linguistic expressions, it always 



appears as an attribute of something else and not as some¬ 
thing independent, we conclude that it is a power. The same 
is true of Time and Inherence. When we say ‘this bamboo 
is straight’, we are giving expression to a cognition which 
does not involve the notion of limit ( avadhi ) and limited 
(avadhimat) and does not depend on something else. It is 
caused by the notion of Direction. The idea of something 
being bent can also be explained by the same notion only. 
It is this very notion which manifests universals like 
bhramanatva’ and ' utksepanatva’ which exist in particular 
actions like ‘turning’ and ‘raising’. This Direction is one, but 
due to particular associations it is looked upon as ten in 
number. That particular region which is in contact with the 
sun daily at the beginning of the day is called the East. The 
West, the North and the South are similarly named. The 
intermediate regions are called by their respective names 
because of the association of the Sun with the regions looked 
after by the Guardians. A particular association with the Sun 
is the auxiliary cause which brings about the apparent divi¬ 
sion of what is One.] 

4. The notion of ‘ first ’ and ‘ next ’ in regard to 
concrete limited objects is based on regional distinction. 
The notion of earlier and later, consisting of sequence, 
arises on account of time. 

[Form (■ murti ) means measurement of things which are 
limited in extent. Things having form are here referred to 
as concrete and limited (murta ). The all-pervasive entities 
such as space ( dkdsa) has no prior-posterior distinction. It 
is about limited objects that we use such expression as ‘first’ 
and ‘next’ on the basis of their contact with the first or the 
next region. Priority or posteriority of a region is based on 
Direction. The notion of ‘earlier’ and ‘later’ amounts to 
sequence which is an effect of the Time-Power of Brahman. 
Thus Direction and Time are two distinct things.] 

It is now pointed out that immaterial things can also have 
artificial distinctions of priority and posteriority. 

ill. 6 . 5 


5. Direction becomes the basis ( apadhi ) for the 
contacts and separations of Space with its divisions and 
of parts of other objects with them. 

[Even all-pervasive entities can have artificial divisions 
such as ‘first’ and ‘next’ or ‘farther’ and ‘nearer’. It has al¬ 
ready been pointed out that space can have artificial divisions 
on the basis of objects which are in contact with it. This 
contact takes place in a particular region ( desa ). That region 
is an artificial division of space. There is contact between 
such regions and parts of the objects which are on them. 
These objects are also described as ‘first’ or ‘next’ and ‘farther’ 
or ‘nearer’ on the basis of such a description of the divisions 
of space. Direction becomes the basis for the contact between 

Remarks . The words Akdsa, Dik and Desa occur in the 
V dkyapadiya. It is, therefore, necessary to try to understand 
the three notions for which they stand. In Vaisesika, Akdsa 
and Dik come under the category of Substance. The former 
is defined as that Substance which is the substratum of the 
quality called sound. It is an all-pervading substance and is 
the medium for the propagation of sound. It is usually 
translated as ‘ether’. It fills the whole of Space and is, there¬ 
fore, different from it. In the Upanisads, however, Akdsa 
often means Space itself and not a substance which fills space. 
In the Vdkyapadiya Akdsa is looked upon as One without 
any real division of its own. In this sense, it is equal to space. 
That is why it is in contact with all the objects of the world 
to which it provides room or accommodation. Helaraja 
actually says that akdsa is so called because it provides room 
(avakdisa) to everything: avakdsaddndddhi taddkdsam. (Vak. 
III. Ja. 15). Even though One, it can be looked upon as many 
in terms of the various objects of the world which occupy 
space. The space occupied by each object is a part or a 
a portion of the one Space. Each of these parts or portions 
is a region (desa) within Space. Moreover, these regions 
have a position in the scheme of things. We can say of a 
particular region that it is to the right of an object or to the 


left of- it. There is something which is the cause of such 
notions as right and left and that is called Dik (Direction). 
According to the Vakyapadiya, it is not a separate entity. 
It is only a power or capacity of what is real. To say, 
t erefore, that Bhartrhari does not make any distinction 
between Akasa and Dik would not be correct. Akasa is Space 
and Dik is position or Direction. In this translation, I have 
used the word Direction for Dik. 

6. Regions are regulated through Direction and 
nothing regulates ‘Direction. Powers of objects are 
postulated on the basis of the service which they 

natule a region is to he the substratum of things. 
Whatever distinction of priority or posteriority there is among 
regions cannot arise out of their own nature. That depends 
upon some other factor and that is called Dik. Thus, we 
postulate the existence of Dik because of the service which 
it renders. Being a power it rests on something else 
(panitantra). That being so, there cannot be another power 
to regulate it. If another power is postulated, Direction (Dik) 
would cease to be ‘power’, because it would become ‘saktimal’, 
the one having power. The quality of being prior, posterior 
etc. is inherent in Dik. Contact with the Sun only manifests 

7. What is called the East is something imposed 
on objects. That due to which the notion of priority 
arises is Direction. Otherwise, it would be a mere 

[What is called East is an attribute imposed on objects. 
It is because of Direction that things are talked of as being 
in the cast or as being ‘prior’. If it is not so regarded, it would 
be a mere name, there being no independent entity like that. 
So it should be looked upon as a power, postulated on the 
basis of its effects.] 

in. 6. 8-9 137 

It is now shown that expressions like prior and posterior 
cannot be based on one’s own body. 

8. The Directions would not be fixed if they were 
based on one’s body. What is behind when one faces 
west would be in front when one faces the other way. 

[It has been said so far that Dik has to be postulated in 
order to explain the notions of the prior, posterior, etc. But 
cannot they be explained in relation to one’s own body? What 
is in front of one’s body would be ‘prior’ (p urva) and what 
is behind would be posterior (p ascdt). We also use such 
expressions as luistadaksina and luistavama (M. Bha. I. p. 118, 
1. 23.) to designate what is on our right or on our left, showing 
that our own body can be the basis of the expression. This 
view is not correct, because there would be no fixity in our 
notions of piirva, para etc., because one would be constantly 
changing the position which one faces. Therefore, it is better 
to accept the view that notions of priority and posteriority 
are based on Dik which has come down from time immemorial.] 

9. The determination of the regions is not based 
on the notion of Directions ( diksu ). One does refer to 
as ‘piirva’ what has been long known as apara. 

[It has been stated that the notions of East, West etc. 
are based on Dik (Direction). Here a difficulty arises. Some¬ 
times we apply the term ‘piirva’ to what from the point of 
view of Dik, is ‘pdscdtya’. When we face west, the region in 
front of us is in the west but we use the expression ‘purva 
for it. When one goes from the South ( Daksinapatha) to the 
Eastern country, one is said to go to ‘purva’, but he is really 
going north ( uitara). Thus, it is clear that we sometimes use 
the expressions purva, para etc. without meaning Dik by 
them. In other words, the notions of ‘purva’ (in front) and 
pascdt (behind) are not based on Dik. The remark of the 
M. Bha. on ‘purvasmin dese vasati’ confirms this. The M. Bha. 
is trying to explain why the suffix ‘ asldli’ cannot come after 
V-II1. 18 


the word ‘purva’ in ‘parvasviin dese vasali \ Tlie reason is 
that the suffix in question is taught after words like purva 
when they directly express desa (region) whereas, in the 
expression in question, the word ‘purva' is only an adjective 
qualifying the word ‘desa’ which is the word expressive of 
region. The word ‘purva’ is not a ‘diksabda’ expressive of 
‘desa’ as required by the sutra P. 5.3.27. Those who maintain 
that the word here specifies the particular direction ( digvi - 
sesdvaccheda) where the region is, are wrong. The conclusion 
is that in applying the words purva etc. to Dik, there is no 
upadhi or underlying notion, because they are rudhi words, 
that is, words based on immemorial usage. In applying the 

words to desa (regions), dik (direction) is the underlying 

It is now shown that this conclusion is supported by the 

M. Bha. 

10. It is because the words ( daksina and uttara ) 
have not the same meaning when they are in the mas¬ 
culine gender that they cannot take the masculine form 
(pumvadbhdva) . In this sense (of dik), the masculine 
gendei ( prasavah ) is never expressed by the words. 

[That is why there cannot be assumption of the mascu¬ 
line form by the words daksind and uttara when the suffix 
‘atasuc’ (P. 5.3.28) is added to them. These two words have 
a masculine form also, but with a difference in meaning. When 
they are masculine, they express ‘vyavastlid’ and not dik. 
Unless the meaning is absolutely the same whether the word 
be masculine or feminine, there cannot be assumption of the 
masculine form (pumvadbhdv a). If it is held that even when 
these words express Direction (dik), the idea of vyavastlid 
(limit, mutual requirement, starting-point) is present, there 
would still be difference between ‘diksabda’ and desasabda .] 

11. Words like purva are invariably used in the 
sense of Direction (Dik). It is like the word sasti (in 

III. 6. 12 


the word sastika ) which denotes time when duration 
of life is measured. 

[The words East (Purva), etc., are applied to the direction 
(Dik ) irrespective of any other implication. But when they 
are applied to the Regions or to the divisions of Time, they 
are based on ‘vyavastha’ (mutual requirement). When applied 
to Directions, these words are really proper names. The notion 
of limit or mutual requirement may be there in a completely 
hidden form. In the expression ‘purvo gliatah’ the word 
‘ghata’, conveys its own meaning and the word ‘purva’ pro¬ 
vokes the question: in relation to what? Thus there is a clear 
difference between purva, etc. as applied to Direction and as 
applied to the Regions and the Divisions of Time. Because 
of resemblance, they appear to be the same. But they cannot 
be the same. What is based on mere immemorial usage cannot 
be the same as something based on a cause. In the case of 
Regions and Time, the words only look like diksabda but are 
really not so. Such words are used in a fixed meaning. It 
is like the word sasti in sastika. Though a numeral, it denotes 
time here without the help of any other word, by mere usage. 
Similarly, purva, etc., denote Directions (Dik) by long usage, 
even though they may denote vyavastha elsewhere. That is 
why they are not ‘bhdsitapuviska ’.] 

12. (It is on account of Direction that) division 
based on light and shade is possible in mountains etc. 
Such a division is not possible in things which do not 
have that attribute. 

[Another effect of what is called Dik is that things like 
mountains seem to be illuminated by the light of the sun on 
one side and to be covered with darkness on another. Without 
Dik this division of parts would not be possible. It serves as 
an indication for the inference of Dik.) 



13. It is through Dik that division is made in the 
indivisible atom. That indeed has been declared to be 
the first power for effecting all division. 

[Division of parts in limited material objects is due to 
Direction (Dik ). Even in atoms which are supposed to be 
indivisible, one can imagine four sides, a top and a bottom on 
the basis of Dik. Other atoms can come into contact with 
these six parts and produce such things as binaries etc. It 
is because of this division into parts from the very beginning, 
that it continues in all the later products. So division of all 
material objects into parts in ihe first service which Dik ren¬ 
ders. Some explain this and the previous verse as follows : — 
‘In all divisible objects, division into parts is due to light 
and shade. In all indivisible objects like the atom, division 
is imaginarv and it is brought about by Dik." What they mean 
is this: Where there is some basis for the division of parts 
such as light and shade as in the case of divisible objects, that 
is the basis of the division. Where there is none as in the 
case of indivisible objects like atoms, there it is Dik which is 
the cause of such division. But all this is wrong. It is due to 
Dik that division in any object is possible, not merely in the 
indivisible atom.] 

14. Objects are (in themselves) without regions, 
divisions, sequence and contingencies. Variation in 
their power takes place because of difference in their 

[Why not assume that things have a special disposition 
of parts in their own nature? Why bring in Dik to explain 
this? The answer is: Things like * dkdsa which are all- 
pervasive have no divisions at all. Material wholes have no 
divisions which are part of their own nature. In both cases, 
parts are artificially made. If they have no parts, there can¬ 
not be any question of sequence of these parts. Because 
of difference in their associations, they seem to have varia- 

III. 6. 15-16 


tion of power and so appear to have divisions and sequence. 
In the same way, when divisions in them are brought about 
by Diky we describe them as prior and posterior.] 

It is now explained how visible material wholes are really 

15. Indivisibility is the same in an atom and in a 
jar. What is called their division is only a power and 
so is their dimension. 

[One might here ask: ‘One does see objects like jars, 
associated with different regions, having visible parts and 
magnitude. How can they be looked upon as indivisible? The 
answer is this: A jar and an atom are alike in that neither 
has any parts in its own nature. Divisions having sequence 
are imagined in them on the basis of the power called Dik. 
These parts are not different from the wholes. Because of 
the relation of inherence ( samavdya )) between the whole and 
the parts, they do not appear to be different from one another. 
If objects had parts in their own nature their unity would 
be destroyed. Division based on association with an exter¬ 
nal thing is not real division. Otherwise, even the atom would 
become divisible. Nor can one say that division is real 
because it is caused by dimension. Dimension is a power 
and it is the cause of the cognition of the ‘smallness’ or 
‘bigness’ of things.] 

16. That by which a division is made is also sus¬ 
ceptible of division. And a division which has no end 
has been said to be improper. 

Remark. The last iicida of this verse, printed c ato 
yuktataram viduh' should really read as e ato’ yuktataram 
viduh. There is elision of the initial ‘a’ of ‘ayuktataram’. 

[No division artificially made through an external factor 
is part of the nature of a thing. An artificially made part 



can also be artificially divided into further parts, until one 
reaches the atom. Even there, one can imagine parts on the 
basis of Dik. Unless Dik itself has parts, it cannot cause parts 
in other things. Divisions in Dik are based on association 
with the Sun. Association with the Sun is based on differences 
in the regions of Mount Meru and that also on something 
else. There would be no end to this process. Division which 
does not come to an end is unacceptable. Thus things are 
indivisible in themselves and they are divided on the basis 
of limiting external factors. All divisions have, therefore, 
only a relative reality.] 

17. Because its effects can be seen everywhere, it 
is said to be all-pervasive. Its all-pervasiveness consists 
just in this. It is otherwise with corporeal things. 

[The effects of Dik, previously explained are found 
everywhere. That is why it is said to be all-pervasive. The 
all pervasiveness of immaterial things consists just in their 
producing effects everywhere, whereas that of material 
corporeal things consists in their occupying extensive space 
with their parts.'] 

18. The assumption of the existence of Direction 
(Dik) and Time (Kola) is as established as that of 
Consciousness. Who indeed would order otherwise 
that which is the very nature of living beings? 

[Even a thing which exists is as good as non-existent if 
it is not cognised, because it cannot fulfil any purpose. 
Consciousness is, therefore, accepted by all disputants. The 
self is of the nature of consciousness. Similarly, Direction 
and Time are based on our cognitions. As explained before, 
things have neither difference nor identity, neither existence 
nor non-existence. They appear in many colours on account 
of beginningless avidyd. All experiences take place in time 
and direction. Otherwise they would not lead to any pur- 

III. 6. 19-20 143 

poseful activity. It is the nature of beings (pralcrtim 
prdnindm ) to have experiences in dik and fcala.] 

19. There would be confusion of activities if this 
nature were not accepted. Therefore, even while 
abandoning these things, one has again recourse to them. 

[If this conception of Dik and Kdla is not accepted, if this 
assumption of Dik and Kdla which has become our nature 
is abandoned, there would be confusion in regard to the 
observance of Vedic injunctions. Even though the Universe 
is without sequence spatially and temporally, still even one 
who knows the truth must act on the basis of the assump¬ 
tion of Dik and Kdla. Even he can cognise things only as 
having sequence.] 

It is now stated that Dik performs a function parallel to 
that of Kdla. 

20. From that Power, distinctions such as East etc. 
arise on the basis of association with other things. 
Thus differentiated, Dik brings about differentiation in 
the cognition of things. 

[Just as Time, the power of Brahman, exercising the 
functions of permission and prevention, brings about tempo¬ 
ral sequence, in the same way Dik is also the cause of the 
notion of spatial sequence among objects, due to association 
with other things. By ‘other things’, contacts with the sun 
are meant. Dik brings about cognitions of objects as quali¬ 
fied by itself. It is then that Dik which is One, appears to 

be many.] 

The result of the notion of Dik in grammar is now set 


21. When the idea of limit (starting point) is to 
be expressed it is the ablative of position which is used. 
When some other idea (as that of part and whole) is 
to be expressed, it is the genitive case which is seen as 

in: ‘the upper part of this’ (purvam asya ). 


[Certain operations in grammar are based on the notion 
of Dik. In the world, there is the idea of limit: ‘this is on 
the east of that’ and so on. There one uses the ablative case. 
Where some other idea such as that of part and whole is to 
be expressed, the genitive case is used according to P. 2.2.1.] 

It is now explained why the notion of prior and posterior 
is not based on one’s own body. 

22. Where there is no confusion in regard to the 
limit, there cannot be any mistake in the use of the ex¬ 
pressions prior (or east) etc. Nor does an indication 
like the expression 'this is straight’ become false. 

[Where there is fixity in regard to the starting point, 
there is no confusion in the meaning of words like prior, 
eastern etc. Notions of prior etc. if based on Dik (Direction) 
are fixed. What is prior to (to the east of) something cannot 
be posterior to (or to the west of) the same thing. Also, 
such an expression as ‘the base of this bamboo is straight’ is 
a clear indication of the notion of Dik and is not liable to 
confusion. It does not depend on any notion of starting- 
point. If it did, there might be occasion for confusion. If 
one has to have a starting point, it is better to take contacts 
of the Sun as such.] 

It is now stated that Dik is something internal, 

23. It is an internal function ( dharma ) which ap¬ 
pears to be external. Or rather, according to this mode 
of thought, there is neither internal nor external. 

III. 6. 24-25 


[According to those who hold that it is an internal 
function, me universe is maniiested witmn me consciousness 
liseu, tnougn on account ot an eternal metaphysical limita¬ 
tion, it appears to be external to it. Dik is in the same posi¬ 
tion. It is something which exists within. There is noiiimg 
corresponding to it outside. In fact, notions of internal ana 
external are relative. In reality, there is neither the one 
nor the other. Notions such as prior and posterior proceed 
from aviclya. There is no such thing as Dik as an external 

It is now stated that it is really useless to discuss whether 
Dik is one or many. 

24. Knowing that the notion of unity and plura¬ 
lity in regard to these powers are speculative and have 
no relation to the thing itself, one must not take them 
to be real. 

[The powers of Dik are primarily one but appear to be 
many due to association. Or they are primarily many, being 
eternal and inferred from their many effects. Both these 
views are only speculations and have no relation to the thing 
itself. One must not take them to be real.] 

25. In regard to objects whose reality is beyond 
speculations, the world follows the usages based upon 

[Things in the world are quite different from what the 
different thinkers ( tirthika ) speculate them to be. They 
obviously cannot have the contradictory characteristics which 
thinkers attribute to them. Therefore, one has to follow the 
usages based on convention. Even thinkers have to do it. 
This applies to the ten Directions (Dik) current in the world.] 

V-11I. 19 


26. There is no unity and no multiplicity and with¬ 
out unity there cannot be multiplicity. In reality, there 
is no difference between the two. 

[If things are mutually dependent, the disproving of one 
would disprove the other. Thus neither multiplicity nor unity 
of Dik would stand examination. Worldly usage is based 
upon what appears to the mind. In reality, both multiplicity 
and unity are unreal.] 

It is now stated that unity and multiplicity cannot be 
predicated of Power. 

27. There is not the same difference between the 
powers as between objects having power. Nor is 
there any worldly unity in their nature. 

[Objects having power can be cognised as something 
distinct. But powers which can only appear as resting and 
dependent upon their substrata, do not appear as distinct 
from one an